“Insulin Dynamics”

This one has a bit for everyone.


Relationship of Insulin Dynamics to Body Composition and Resting Energy Expenditure Following Weight Loss (Hron et al., 2015)


I think study was actually done a few years ago, originally published here (blogged about here), and re-analyzed through the eyes of Chris Gardner.  I think. (But it doesn’t really matter as the study design appears to be identical.)


Experiment: give someone an oral glucose tolerance test (75 grams glucose) and measure insulin 30 minutes later.  Some people secrete more insulin than others (a marker of insulin resistance); these people also have a lower metabolic rate after weight loss = increased propensity for weight regain.  However, if these people follow a low carbohydrate diet, then the reduction in metabolic rate is attenuated.  Some people who don’t secrete a lot of insulin after a glucose load may do better in the long-run with a lower fat diet.



It’s an extension of the basic insulin theory: more insulin secretion = less nutrients available in the blood, and if food intake is strongly controlled (like it was in this study), energy expenditure declines to compensate.

Disclaimer: different study design, different paradigm, different result.


There are two relevant figures in this paper.  Here’s one of ‘em:



insulin secretion and fat mass



If insulin is low 30 minutes after a glucose load prior to the weight loss intervention, then more fat mass is lost during said intervention (P=0.04).  Regardless of diet composition (pretty much).

“At the same amount of weight loss, individuals with high insulin response lose relatively more lean mass and less fat mass than those with low insulin response.”



In the original study, everyone lost weight via calorie restriction, then were randomized to isocaloric low, medium, or high carb diets for weight loss maintenance.  Then switched, and switched again; a three-way crossover.


Here’s the rub: the low carb diet reduced the insulin response to a glucose load (ie, improved insulin sensitivity), and this effect is maintained for a while after carbs are re-introduced.  So in the crossover design, if you were assigned to a low carb diet first, then your insulin response to glucose is still lower on the next diet, and as it goes, energy expenditure is still improved.

The magical ~300 kilocalories is back again.  Figure 2:


effect of first diet on energy expenditure


[aside: check out the light grey line all the way to the left; these are the insulin sensitive people who did better on high carb.  The diet-difference is, however, most profound for insulin resistant people (compare light grey and dark grey lines all the way to the right).]


This is a bollixed crossover (in a sense), and what we’re seeing is an “order effect” because the same thing doesn’t happen to those assigned to a high carb diet first… but it’s an interesting angle of the insulin hypothesis!  [You can probably interpret this another way (I certainly can), but I rather like this way.]





order effect


…and according to this analysis, reduced insulin secretion is associated with higher energy expenditure, and better chances for sustained weight loss (theoretically*).


Energy expenditure


*we know that in real life, diet alone rarely works.  Lifestyle interventions which improve circadian biology are often required, as we’ve seen studies showing artificial light regulates fat mass (in rodent models, but probably humans, too), sleep quality has a major impact on body composition, etc., etc.  A disrupted circadian rhythm puts you in the weight-reduced, energy conserving state before you’ve even lost weight.


No, the insulin hypothesis is not dead, but it’s not the whole story.


Another important disclaimer/confounder: weight loss per se (regardless of method) improves insulin sensitivity and overall health in many people.  Best advice is to find the method which is the least painful for you, that works.  At the end of the day, sustainability & adherence = 100%.


Also, read this article by Dr. Corkey, and don’t be an insulin hyper-secretor: Banting Lecture 2011: hyperinsulinemia, cause or consequence?

For full access to all articles and much more (or if you just like what I do and want to support it), become a Patron! Five bucks a month for full access and there are many other options. It’s ad-free and you can cancel if it sucks ????

Also, I’m open to suggestions so feel free to leave a comment or email me directly at drlagakos@gmail.com.

Affiliate discounts: if you’re still looking for a pair of hot blue blockers, Carbonshade is offering 15% off with the coupon code LAGAKOS and Spectra479 is offering 15% off HERETrueDark is running a pretty big sale HEREIf you have no idea what I’m talking about, read this then this.

20% off some delish stocks and broths from Kettle and Fire HERE

If you want the benefits of  ‘shrooms but don’t like eating them, Real Mushrooms makes great extracts. 10% off with coupon code LAGAKOS. I recommend Lion’s Mane for the brain and Reishi for everything else.

calories proper


Be Sociable, Share!
  • Jan

    Not sure if entirely relevant to this post, but curious anyway. Tentolouris (2003) found, amongst other things, a much greater individual variation to high carb vs high fat isocaloric meals in both lean and obese women.

    Are these findings consistent with other studies done on this, and are they a fair representation of the individual variation in response to high carb va high fat isocaloric meals in a general population?

    Any suggestions/links, please?


  • valerie

    I’d be curious to see a sensitivity analysis on this dataset. From the first graph, it seems that the trend is entirely driven by the three most insulin resistant subjects. Remove those, and maybe the conclusion would be quite different (like, diet composition doesn’t matter unless you are in the top 20% insulin resistant, for example).

    Do you have access to the whole dataset? It’s only 21 subjects, after all.

    • “the whole dataset” is shown in the figure.

      There are also three subjects in the top left; removing them could cancel out the effects of removing the three in the top right.

      But more importantly…

      “If insulin is low 30 minutes after a glucose load prior to the weight loss intervention, then more fat mass is lost during said intervention (P=0.04). Regardless of diet composition (pretty much).”

      in other words, this graph is from the ‘weight loss’ phase of the study (calorie restriction), prior to the low/med/high carb weight loss maintenance phase.

      • valerie

        “the whole dataset” is shown in the figure.

        No. The figure, as you say, shows only the weight loss phase. I mean the WHOLE dataset: baseline weight and insulin, weight lost, plus REE and weight change during each of the maintenance diets. For each of the 21 subjects, not averaged.

        I don’t know how you can infer anything about amount of weight lost vs diet composition in this study. Didn’t they have only one diet intervention during the weight loss phase?

        My hunch (and it is just a hunch) is that diet composition does not matter for weight loss or REE for anyone except the very most insulin resistant individuals. In this case, 3 of the 21 subjects. But I would need the whole dataset to verify that.

        In other studies you posted about previously, the authors had often used the median as a cutoff. The least insulin resistant half of the subject pool (most of whom were probably still somewhat insulin resistant since they were obese) did not benefit from a low-carb diet. It leads me to suspect that the level of insulin resistance required to do well on a low-carb diet is rather high.

        • while I like the idea of dividing at the median, I think top 20% vs. bottom 20% would only magnify the difference (unless you were referring to top 20% vs. bottom 80%?).

          also, I think we disagree on this point: “…the level of insulin resistance required to do well on a low-carb diet is rather high”


          • valerie

            I meant top 20% vs bottom 80% (of the entire population). That would include all people with type 2 diabetes, a tiny portion of lean people, some overweight people, and a part only of obese people (maybe close to half of them, maybe quite a bit less, I am just hypothesizing here).

            Now, let’s say that a low-carb diet works for that top 20% insulin resistant people. Other people with obesity are left basically with no efficient intervention to lose weight. Old tricks like portion control (calorie restriction) and lower calorie density ( low fat) would be their only tools. Told to eat as much as they want of a low-carb diet, they don’t lose an ounce, because their main problem was never carbs in the first place.

          • Insulin sensitive obese patients can still lose a considerable amount of weight on low carb diets (see studies by Pittas and Cornier), and in some cases, just as much as they would on a low fat diet (see study by Ebbeling).


          • valerie

            Pittas and Cornier = calorie restricted
            Ebbeling = only insulin resistant people on a low-carb diet (unrestricted calories) had good results

  • Is this study really just a re-analysis of the 2012 study by Ebbeling ? 21 people, 3 way cross over etc..

    • yes, I’m also positive (even the baseline patient characteristics are the same).

      • Thanks for blogging on this – it’s really fascinating stuff.

  • Thomas Hemming Larsen

    You say that weight loss is good for improving insulin sensitivity. Wouldn’t increases in muscle mass have the same effect? More muscle to soak up the glucose.

    • Yes. Also, the exercise you presumably did to increase muscle mass directly clears blood glucose via muscle contraction.

    • Sky King

      Bodybuilding 101. Hence why men in general can eat more carbs w/o gaining fat than woman can. Men have more muscle mass than do women.

  • rs711

    Hi Bill,

    Can you expand on why “Some people who don’t secrete a lot of insulin after a glucose load may do better in the long-run with a lower fat diet”?

    Scenario 1: you do not secrete a lot of insulin, but are quite IR, & so have high blood glucose levels circulating longer.

    Scenario 2 : you do not secrete a lot of insulin, are quite IS, & so have lower blood glucose levels & the peaks don’t last for long.

    In scenario 1 it seems the LF diet would be counter-productive.
    In scenario 2 it seem the LF diet would be less problematic (or potentially OK).


    • scenario 1: this is late stage T2DM (a population that hasn’t been studied through Gardner’s eyes); they’d likely fare better w/ LC.

      scenario 2: I might’ve guessed they’d do well on either diet, but studies show better weight loss with LF. Since insulin sensitive, no hyperglycemia on either diet.

      [does this answer your question?]

      • rs711

        yeah you answered it, thanks.

        [The basis of our disagreement seems to circle back to what we take away from the Pittas, Cornier & Ebbeling references in http://caloriesproper.com/insulin-resistance/. No point in resurrecting that dead horse now; maybe I’ll address it more completely at some point.]

        Its fine to look at the ‘absence’ of hyperglycemia in scenario 2 but i dont think it’s the end of the line….What what happens to a cells’ NADH:NAD+ ratio? Its ROS levels? I honestly can’t think of a scenario where a LF scenario does well there :/

  • Hi Bill,
    IMO, finding the method that works is 90%. Sustainability & adherence = 10%. At the end of the day, unless the method works you get nothing from it.

    RMR is easier to measure, but it is not TEE.

    • rs711

      1) “adherence” = the result of doing something that ‘works’ (whether that’s positive or negative – people do adhere to vegan diets for years & years after all..)
      2) “sustainability” = a longer-term (indefinite?) result of doing something that ‘works’

  • Esmée La Fleur

    Another great post!

  • Eve
    • interesting. Thanks!

      • Joe Gavin
      • Eve

        Bill, just out of curiosity, what protein intake would you consider optimal for a female weightlifter? Additionally, how would you reconcile the logic underlying Peter’s diet, and its emphasis on minimal/adequate protein, with diets that promote >adequate protein for aesthetic or functional purposes? Finally, what macros do you consume, if you don’t mind my asking? Thanks in advance!!!

  • George

    “However, if these people follow a low carbohydrate diet, then the reduction in metabolic rate is attenuated.”

    John Yudkin said this in This Slimming Business – pull out fats to lose weight and BMR can go down, pull out carbs it doesn’t.