Glucagon, Dietary Protein, and Low-Carbohydrate Diets

Glucagon is a hormone that plays an important role in blood glucose control.  Like insulin, it's secreted by the pancreas, though it's secreted by a different cell population than insulin (alpha vs. beta cells).  In some ways, glucagon opposes insulin.  However, the role of glucagon in metabolism is frequently misunderstood in diet-health circles.

The liver normally stores glucose in the form of glycogen and releases it into the bloodstream as needed.  It can also manufacture glucose from glycerol, lactate, and certain amino acids.  Glucagon's main job is to keep blood glucose from dipping too low by making sure the liver releases enough glucose.  There are a few situations where this is particularly important:

  1. Hypoglycemia.  When blood sugar drops below a certain threshold, for example if a diabetic injects too much insulin, the brain senses it and initiates a response (the counterregulatory response) to bring glucose back up and prevent unconsciousness and death.  Glucagon release is an important part of this response.
  2. Fasting.  Glucagon helps support blood glucose levels during fasting, when glucose intake is zero, by stimulating the production and release of glucose by the liver.  This sustains the brain, which has an absolute requirement for glucose (though it can derive some energy from ketones).
  3. High-protein meals.  Protein stimulates insulin release as much as carbohydrate does (because one of insulin's jobs is to send amino acids into lean tissues such as muscle), but protein doesn't supply rapid glucose like carbohydrate does.  If this process went unchecked, eating a high-protein meal would cause hypoglycemia because insulin release would suppress blood glucose too much.  Glucagon release counterbalances insulin, preventing hypoglycemia when we eat a high-protein meal.
This post will discuss the third scenario.  If we're looking to reduce circulating insulin levels using diet, which is ostensibly the goal of certain low-carbohydrate diet strategies, reducing carbohydrate intake is only half the story.  If we really want to eat low-insulin foods, the first thing to do is look up which foods actually elevate insulin following a meal.  The data, published in a paper titled "An Insulin Index of Foods", may surprise you.  This paper is free access.

What we see right away is that high-protein foods frequently stimulate insulin to a similar, sometimes even greater, degree than high-carbohydrate foods, calorie for calorie.  Beef and fish release as much insulin as brown rice.  Pasta (white or brown) and porridge release less insulin per calorie than cheese, beef and fish.  Donuts are only 25 percent higher than fish.  According to the paper, the five lowest-insulin foods tested were (from lowest to highest) peanuts, eggs, bran cereal, white or brown pasta, and grain porridge.  Doesn't sound very low carb, does it?

People frequently cite glucagon to resolve this problem.  The idea goes like this: glucagon is the opposite of insulin, and if they're released together, as they are when you eat a high-protein meal, then their effects on blood sugar, on hunger, and on fat metabolism cancel one another out in a way that they would not following a carbohydrate-heavy meal.  If we're talking about blood glucose, this is correct.  Glucagon does more or less cancel out insulin's effect on blood glucose, and eating protein in isolation does not lead to major changes in blood glucose.

Regarding hunger, glucagon is often suggested to oppose the hunger-inducing effects of insulin.  However, contrary to popular claims, insulin doesn't increase hunger or food intake in humans unless it causes frank hypoglycemia, so there is nothing to oppose (1, 2, 3, 4)*.  Nevertheless, glucagon probably does play a role in satiety, independently of insulin.  So that claim is partially true.

If we consider the claim that glucagon promotes fat release from fat tissue, suddenly we're on shaky ground!  When researchers put high doses of glucagon on fat cells in a petri dish, or give very high doses of glucagon to animals or humans, it stimulates the release of fat (lipolysis).  So the idea that glucagon counterbalances insulin's effects on lipolysis does have some basis in reality.  But giving humans realistic doses of glucagon, doses that approximate what would occur naturally in the human body following a high-protein meal, does not increase lipolysis (5, 6).  Also, blocking glucagon action in dogs does not reduce lipolysis, suggesting that baseline glucagon levels are unrelated to lipolysis (7).  The glucagon receptor knockout mouse is actually lean and resistant to diet-induced obesity, contrary to what these claims would predict (8, 9)**.  That's why modern reviews, such as Keith Frayn's textbook Metabolic Regulation, make statements such as these:
Glucagon has a potent effect in isolated fat cells in the laboratory, but appears not to affect fat mobilization in humans in vivo.
Glucagon probably does play a role in satiety (fullness).  As with many other satiety peptides released by the digestive tract and pancreas, glucagon is sensed by nerve terminals that send impulses up the vagus nerve, and the signal is transmitted to the brain where satiety is perceived (10).  In the case of glucagon, the relevant receptors are located in the liver, where glucagon mediates the majority of its effects***.  Injecting glucagon decreases food intake, and blocking glucagon action increases food intake, though the evidence has not always been consistent (11).  Protein, and to a lesser extent alcohol and fat, but not carbohydrate, lead to increased glucagon secretion following a meal (12, 13, 14).  Increased glucagon secretion may be one reason why high-protein diets lead to reduced food intake and body fatness in overweight people, although other convincing mechanisms have been proposed****.  The body's response to food is so complex that it's tough to predict large-scale (and long-term) physiological changes by measuring a single hormone, but it seems likely that glucagon plays a role.

Whatever the mechanism, protein remains the most satiating macronutrient, and it does help with fat loss.  In fact, recent evidence suggests that the ability of low-carbohydrate diets to promote fat loss and maintenance (relative to low-fat diets) may have more to do with increased protein than decreased carbohydrate (15), at least at a moderate level of carbohydrate restriction.  It is known that high-protein, high-carbohydrate, low-fat diets are effective for fat loss, and this diet stimulates a large amount of insulin release (16, 17).

In one particularly interesting study, increasing protein intake at the expense of fat (with no change in carbohydrate percentage) led to a spontaneously reduced calorie intake and substantial fat loss that was comparable or superior to what is typically observed in low-carbohydrate diet studies (18), particularly considering that the participants were only modestly overweight at baseline and were not even trying to lose weight.  Both fasting and post-meal insulin levels remained unchanged throughout the intervention.  None of these effects seem to involve the supposed influence of glucagon on fat cells, or anything related to insulin, although they could relate to the effects of glucagon on satiety.

Together, this paints a complex picture, suggesting that the effectiveness of low-carbohydrate diets for fat loss in overweight people:
  1. Depends at least in part on increased protein consumption.
  2. Probably does not require a reduction of insulin secretion, but may involve an increase of glucagon.
  3. Probably does not involve direct effects of glucagon on fat cells, but could relate to glucagon's effects on satiety, and perhaps other effects in the brain.
Beyond the metabolic/hormonal effects of low-carbohydrate or high-protein diets, these diets may also be effective because of changes in the reward and palatability value of the diet (and food variety, a related factor).  Getting rid of pastries, donuts, soda, fries, and pizza goes a long way, and taking one or more items off the table (e.g. bread) tends to reduce total food intake even if everything else remains the same.  The potato-only diet is a good example of this (19, 20).  Over on Mark Sisson's forum, a number of people have been repeating Chris Voigt's experiment of eating nothing but potatoes, with similar weight loss results (21).  To be clear, I'm not recommending this approach, but it certainly illustrates the power of a monotonous diet.  Interestingly, besides being extremely filling per calorie, potatoes are also one of the most insulin-stimulating foods, according to the paper cited above.


* I am aware of one older study where increasing insulin did lead to hunger and increased food intake independently of hypoglycemia.  That paper had the least convincing study design, but I included it for completeness.  It's the last of the four references I provided.

** They also have much higher levels of GLP-1, which may be a compensatory adaptation, and probably plays a role in the overall phenotype.  GLP-1 is an incretin hormone as well as a satiety hormone.  Incretins are a group of peptides that stimulate insulin secretion.  This is the problem with knockout mice in general-- when you get rid of something completely from birth, it can trigger compensatory adaptations that make the resulting phenotype difficult to interpret.

*** When glucagon is secreted, it goes to the liver first.  Glucagon concentrations are much higher in the hepatic circulation than in the general circulation, which is one of the reasons why it's thought to act primarily in the liver.  However, Dr. Tony Lam's lab has shown that it also acts directly in the brain to regulate blood glucose.

**** The amino acid leucine, increased on high-protein diets, acts in the hypothalamus on mTOR and AMPK pathways, which regulate energy homeostasis (21).  AMPK and mTOR are the cell's primary energy sensing pathways, sort of like leptin and insulin are the primary energy sensors on the organism level.  This mechanism seems to be at least partially responsible for the ability of high-protein diets to improve body composition (lower food intake, lower fat mass, and higher muscle mass) in rodent models.