Response to Insulin Series Part 1

This write up is in response to the first page of the series here.

Insulin is an anabolic hormone with its primary role to stimulate growth and storage. Its roles are complicated varied and we still work to understand the scope of the particular functions of insulin in their entirety. However, we do know unequivocally that insulin helps drive nutrients into cells by promoting uptake and storage of nutrients. Insulin also stimulates protein synthesis, formation of triglycerides, fat uptake in adipose tissue, and glycogen production. Too little or absolute insulin deficiency leads to unrestrained liver glucose production, lipolysis, ketoacidosis and an untimely death. Too much insulin and we experience acute hypoglycemia, dizziness, brain fog, irritability, tissue resistance, metabolic syndrome (diabetes, obesity, hypertension, cardiovascular diease), brain failure, and again death. Lately, insulin signaling has also been implicated in formation of brain plaques in connection with dementia and Alzheimer’s (here) and has been study for years in the acceleration of malignant tissue (Temin, 1969, Osborne et al., 1976). It’s not a coincidence that illnesses such as obesity, heart disease, and diabetes are so commonly manifest together, rather than independently. They have an associative hormonal cause, which is the overproduction of insulin.

Naturally, we can see that insulin is a critically important hormone, but that abnormal secretion, especially chronic, can have dire consequences. To say that insulin is “not the bad guy” without putting any qualifiers behind that statement is just plain irresponsible, especially as a diet and health authority with considerable clout. Insulin is not bad in that it is vital to life, but chronic insulin production from an insulemic diet is a major factor in the current epidemic of metabolic diseases, especially in affluent nations and especially in the last 50-60 years. I suppose I may agree that outside of autoimmunities, insulin alone is not the primary cause of metabolic syndromes (that I would bestow upon the diminished endogenous regulation of blood glucose in favor of exogenous sources – or over consumption of carbohydrates), but it is the disruption of tight insulin signaling that promotes a cascade of disease and aging pathways (see here and here).

And now to respond to specific points made in the insulin series. I will be addressing the author specifically in many of my arguments:

High carbohydrate diet -> high (chronic) insulin -> increased lipogensis and decreased lipolysis.

Yes, contrary to Krieger’s assertion, this is still true. This can then lead to high body fat and obesity. It does not result in same effect in everyone since different people have different insulin sensitivities and levels of secretion as well as different metabolisms. The assertion that insulin is only elevated during time just after a meal is slightly misleading. The model he imposed assumed a regimented and controlled diet of exactly three meals per day without snacking, and with very exacting intervals. If I may say so myself, this looks too controlled and does not fit with the variability and randomness that is day to day human life. Take that statement or leave it, I know you won’t find it broadcast overtly in any study but to ignore it is to compartmentalize human existence into machine-like behaviors, which is far from the truth of organic stochastic patterns. It also does not account for the adaptation the body makes with insulin secretion on a meal to meal basis.

First, high carb meals have a limited satiation value. Krieger himself provided plenty of support and literature for this actually, so I don’t think it needs repeating in exhaustive detail, but here’s a few studies (Bertenshaw et al. 2007, Claessens et al. 2009, Latner and Schwartz 1999, Johnstone et a. 2008). I could go on like this. A high carb meal (I call high carb 40% or greater of calories from carbohydrates, however they are usually in excess of 60% calories) promotes high insulin to glucagon ratio (3.0 – 5.0) versus closer to 1:1. The insulin secreted to deal with the carbs is much in excess of the glucagon stimulated. Insulin decreases blood glucose levels and inhibits catabolism of glycogen stores to restore some free glucose in the blood. So if we establish that high carb diets are not very satiating, then we might predict the next step, which is food craving, between meal snacking, and more carbohydrate consumption. To add a little insult to injury (though I know this was not explicitly mentioned in the Krieger series), the conventional diet recommendations for weight loss are almost always to eat 6 small meals per day, so there is an obvious disconnect between the fed and fasted model in the Krieger article, and the actual meal frequency advice for weight loss dieters. I actually think the 6 meals a day recommendation extends to all people according to weight-loss and management programs, though I believe it is folly and not a natural eating pattern. The reasoning behind 6 meals/day is to prevent insulin spiking, rather, keeping insulin secretion constant and chronic to “stave off hunger and overeating”, though it still requires constant snacking. So much for reducing appetite. Chronic insulin secretion down-regulates insulin receptor expression (the cells reduce the number of insulin receptors on their membranes), and promotes insulin resistance, thus actually leading to high insulin spikes because the pancreas has to release even more insulin than normal to elicit an adequate response from the cells.

Back to the point, Krieger “dismantled” the “myth” that high carb diets induce chronic high insulin by a diagram illustrating regimented 3 meal/day structure and stating that there is enough fasting time with presumably low enough insulin to induce lipolysis. For one, however, most people snack in between meals, especially on a high carb diet (since carbs make you hungry, a point demonstrated in controlled diet trial studies). Two, the insulin response to food is an adaptation based on meal experience over time, so a carb-heavy diet will promote a higher insulin response with each meal, and just as insulin response is adaptive, so is fat utilization in fasted states, with greater use of the fatty acid metabolism pathway in a low carb adapted metabolism. The article Krieger linked to the Okinawa diet actually displays the benefits of caloric restriction (CR), not high carb, though even the authors of this article admit caveats:

“…the Okinawan mortality advantage has all but disappeared except in older cohorts (aged 65-plus), it would be informative to have a more detailed, population-based epidemiologic analysis of the traditional diet, energy intake, energy expenditure, phenotype, and the subsequent mortality experience of this older cohort.”

The conclusion that the Okinawan diet is healthful because it is high-carb (or even in spite of high-carb) is misleading at best, but also an example of confirmation bias. Compare apples to apples and you find that the low carb diets trump high carb diets every time with regard to weight loss, lipid profiles, and energy (Hession et al. 2008, Shah et al. 2007, Scientific American), with high carb and caloric restriction being un-maintainable for most people. I attempted to find the second article Krieger linked, which was about the Hawaii Diet promoting weight loss with high carb, but was unable to access it through multiple different search engines and journal subscriptions (surprising actually, it must be quite obscure).

To the argument against carbohydrates driving fat storage

Krieger stated that the carbohydrate hypothesis of fat storage via insulin is a misconception because the body can create and store fat without high levels of insulin. First, his argument doesn’t really address why he finds carbohydrate driven fat storage to be wrong in particular, so it is sort of a straw-man approach. Regardless, I’ll discuss his argument. True, fat deposition at low insulin level is possible, though this is not in the absence of insulin, since that is the type I diabetic who is completely incapable of retaining body fat. Is insulin needed? To promote fat storage, yes. In high quantities? No. Since I argue that healthy insulin levels are low insulin levels, it behooves me to indeed confirm that the body can create and deposit fat with low insulin (otherwise you’re no better off than a type I diabetic). However, this does not somehow disprove or refute the pathway by which carbohydrates promote fat storage. Also, here’s the key part in the abstract he linked:

“There was increased escape of LPL-derived fatty acids into the circulation from adipose tissue, shown by lack of reesterification of fatty acids.”

The body does not hang on to fatty acids in adipose tissue without high insulin. Sure it will store it, but only briefly, until the fatty acids are released back into the blood stream, either to be used in gluconeogenesis (glycerol) or in beta oxidation to form Acetyl CoA. Carbohydrates provide not only the constituent for triglycerides (alpha glycerol phosphate), but the signaling for triglyceride production via esterification of the glycerol with the fatty acids. Triglycerides are too large to move freely in and out of the fat tissue and require HSL to hydrolyze the ester into a fatty acid. Insulin inhibits HSL and signals triglyceride production in adipose tissue.

“This means you will be unable to lose fat even when carbohydrate intake is low, if you are overeating on calories.”

Not quite. You may not lose weight since you have an over adequate supply of calories, but you will certainly be mobilizing fat. Also, “overeating on calories” is rather vague. There is a metabolic advantage of eating low-carb/high protein versus high-carb/low protein, and besides, invoking thermodynamics to explain weight gain is a bit dodgy itself (see my post on this here).

Insulin makes you hungry/ insulin suppresses appetite argument

Insulin of course would signal for satiation since it is a sign that one has just consumed a meal. However, this signal is crossed with others that can either enhance the satiation signal, or override it. Also, while insulin will signal satiation, the reduction of blood glucose levels insulin will cause without a counter effect from glucagon stimulation will then actually increase appetite. Neurons in the basal hypothalamus both receive and feed forward signals arising from tissues such as the gut, blood, and fat that convey information about the relative availability of metabolic substrates. Glucose availability in the blood is one of these, and its absence stimulates hunger. What drastically reduces glucose concentration? Insulin. So while insulin is one of the peptides that sends an anorexigenic signal (inhibits food intake), insulin’s consequential effect on (lowering) blood glucose sends an orexigenic signal (stimulates food intake). In fact, suppression of insulin reduced carbohydrate cravings and appetite (Velasquez-Mieyer and Cowan 2003)

Protein as a stimulator of insulin

Yes protein, especially BCAA’s such as leucine, stimulate insulin secretion because BCAA’s bypass the liver to go into the blood stream and are then sent to muscle tissue to stimulate protein synthesis. Insulin is needed to direct the amino acids into the cell and stimulate protein synthesis, and so long as insulin is released in pulses, and the nutrients are entering insulin sensitive cells, then this is healthy. Furthermore, only low levels of insulin are really necessary to stimulate muscle protein synthesis (Deldicque et al. 2005). Carbohydrates shut down protein synthesis, and insulin response from carbohydrates promotes triglyceride formation in adipose tissue (Keller et al. 2003).

There is no argument here as to whether protein stimulates insulin, but what I do criticize is the reliance on the Boelsma et al. 2010 article on measurements of postprandial effects after low-carb/high-protein (LC/HP) and high-carb/low-prot (HC/LP) meals. This is another instance I’ll call confirmation bias, or perhaps just a jump to conclusions that are not implicit in the article. That sounds nit-picky, but it is very deceptive to over-interpret equivocal results and akin to snake-oil to give out false conclusions. In the Boelsma study, both groups show nearly the exact same response to insulin. The results are non-significant. The LC/HP group was given equivalent carbs and protein (~75g and 35% of calories). In no way is this low carb and the fact that variables are not isolated only nullifies the ability to conclude anything useful from this study regarding differential effects of macronutrients on postprandial signals. In fact, part of the amino acid signal (glucagon) is partially or completely abolished (the ratio is drastically skewed towards insulin, nullifying glucagon signal activation) even with slight simultaneous carbohydrate consumption, and this study has equivalent carbohydrate consumption (Mueller-Wieland et al. 1970). Seventy-five grams of carbohydrates are roughly 3/4’s of my daily intake, much less my intake for a single meal! (I don’t want my argument to hinge on the anecdotal, but that’s a little perspective at least). I think it’s fair to say that the Boelsma study crosses its wires a bit.

Now, here is why carbohydrates are implicated as singularly driving insulin, and the key point Krieger missed. Carbohydrates increase the insulin/glucagon ratio, meaning more insulin is released than glucagon. Consumption of protein, regardless of its insulin stimulating properties, produces a concurrent glucagon release. Glucagon serves to prevent hypoglycemia induced by insulin and promote insulin sensitivity and reducing the need for chronic or elevated insulin secretion postprandially. Carbohydrates do not do this and inhibit stimulation of glucagon. Insulin released in pulses is fine, chronic insulin release is not. Carbohydrates induce chronic insulin release. It is the diet as a whole over time that matters (Muller et al. 1971).

Yet another problem is that again Krieger falsely concludes from Boelsma study that because the LC/HP meal produced the highest satiation values and lowest hunger score over time, and because there was a coincident insulin response with the HP/LC feeding, that insulin caused the satiation! (I believe he used the term “trend”, obviously because values only approach significance but in fact are non-significant). For shame. How much more blatant confirmation bias can you get? Not only that, but disingenuous conclusions spouted in the context of a serious and scientific discussion obfuscate the truth to casual readers. Both meals caused and insulin response. Neither insulin response significantly deviated from the other and neither postprandial well-being report (including satiation and hunger) significantly differed. These results are equivocal, yet he drew conclusions as if this somehow proves insulin drives appetite suppression. Even if satiation factors were significant in the direction they trended (which was that the LC/HP diet trended towards increased satiation), one absolutely cannot credit these effects to insulin. The study provides ambiguous results on those grounds, however I can provide studies that show protein consumption to be the satiation factor, not insulin (Batterham et al. 2006, Layman and Baum 2004). Why does protein induce satiation? Potentially a few reasons. It triggers tryptophan conversion into serotonin which promotes satiation signals in the brain (Garattini et al. 1988); it promotes glucagon release, which keeps blood sugar levels from plummeting; it spares glucose use by stimulation of the glucose-alanine cycle, and inhibits insulin stimulated glucose oxidation in fat cells and incorporation into triglycerides whereby glucose can enter and feed the brain (Mizunuma et al. 1981).

Beyond the Boelsma study, Krieger referenced a study showing insulin response to 4 different sources of protein (which he mislabeled numerous times, interestingly, as “4 different types of protein.” They are different sources, not necessarily types.). Not surprisingly, the leucine bomb that is whey protein elicits the highest insulin response. Again Krieger attributed the satiation value to insulin (without so much as an acknowledgment to the satiation factors known to protein). You can say that there was a high correlation between insulin and appetite suppression until you’re blue in the face, but that correlation does not turn into causation, especially not when insulin is induced by the appearance of amino acids in the blood.

This concludes the first part of my critique, which has only addressed part 1 of the insulin series. I never expected to go on this long, but I’m hoping a thorough review of part 1 can allow me to get through parts 2-5 a little quicker since much of this information is cumulative.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s