Obesity Is Always and Everywhere an Insulin Phenomenon

Link to the Amazon page for The Obesity Code   Note the cover's tagline "Why your body's own insulin is the key to controlling your weight."

Link to the Amazon page for The Obesity Code

Note the cover's tagline "Why your body's own insulin is the key to controlling your weight."

Milton Friedman's dictum "Inflation is always and everywhere a monetary phenomenon" has some complex ragged edges (in particular the variability of velocity), but it is an excellent place to start if you want to understand inflation and how to engineer disinflation. Similarly, the statement "Obesity is always and everywhere an insulin phenomenon" is an excellent place to start if you want to understand obesity and how to engineer weight loss.

The upward trend of obesity is well-known; you can see some scary graphs in the slides for my talk "Restoring American Growth." In "Restoring American Growth" I point out the importance of the issue by this thought experiment: think how much better we would consider the situation in America today if the distribution of income had evolved in exactly the same way it actually did, but there had been no rise in obesity. 

I view the continuing rise of obesity around the world a first-class scientific mystery. Scholars in many different disciplines, including economics, should take a crack at deepening our understanding of it. This post gives what I think is some of the relevant theory on the biological side.

My understanding of the function of insulin relies heavily on Jason Fung's excellent and persuasive book The Obesity Code, which I highly recommend. Jason discusses available evidence at length and fills in with his own judgment where evidence is lacking. I am not aware of any point at which evidence decisively shows Jason's views to be wrong, though there are definitely areas in which he goes against a conventional wisdom founded on underpowered studies. 

Jason's other book, The Complete Guide to Fasting: Heal Your Body Through Intermittent, Alternate-Day, and ExtendedIn line with his usage, from here on, I will use the term "fast" to describe any period of zero food consumption, while drinking only water, tea, black coffee, or the bone broth Jason recommends as an option. However, I will mainly apply the term to twelve hours or more of zero food consumption.


Insulin, Glucagon and Glycogen

Insulin is a central regulatory hormone for our bodies' handling of food energy. Among other roles of insulin, insulin has these key roles: 

  • High insulin levels instruct muscle cells and fat cells to take glucose (blood sugar) from the bloodstream. In the case of fat cells, the glucose taken up is made into triglycerides—precursors of body fat.

  • Low insulin levels instruct fat cells to convert body fat into glucose and add it to the glucose circulating in the bloodstream.

The current Wikipeda article on "Glycogen" (see below) explains that "Glucagon, another hormone produced by the pancreas, in many respects serves as a countersignal to insulin." So when I write about insulin levels being low, that is a shorthand for insulin levels being low and glucagon levels being elevated.

Glycogen is a molecule made from hooking glucose molecules together in a fairly simple way that is used to store glucose in the liver and muscles. Not all of the effect of high and low insulin levels on energy stores fall on body fat; especially at first, much of the effect falls on glycogen.  It seems likely that as the stores of glycogen get further from the target buffer stock of glycogen (or in the extreme, close to zero or to capacity) more and more of the effects of high or low insulin on energy stores falls on body fat.

To get an idea about the relative importance of fat burning versus glycogen burning, here is a useful passage from the Poliquin Group nutrition article "Eleven Myths About How the Body Burns Fat." (I interpret "carb burning" as "glycogen burning.") 

Fat is your body's primary fuel when you haven't just eaten and aren't exercising (a state we call "at rest"). At rest, when you're sitting at your desk or even going for a walk, your body is burning mostly fat, but depending on the need for energy (such as if you had to walk up three flights of stairs), [your] body may increase carb burning. ...

Here's an example: If you burn 200 calories per hour walking slowly, 60 percent is fat (120 calories) and 40 percent is carbs (80 calories). But if you burn 600 calories running at a pace of 8 minutes a mile, 40 percent is fat (240 calories) and 60 percent is carbs (360 calories). 


The Central Analogy

My central analogy is this:

  • High insulin levels are like a high growth rate of the money supply.

  • Low insulin levels are like a low growth rate of the money supply.

  • Body weight is like the level of inflation.

(However, insulin levels are strongly influenced by the types and amounts of food eaten, so insulin levels are more like endogenous than exogenous movements in the money supply.)

There is one other important piece of the analogy:

  • Fasting (a period of no food) is like a recession.


Executive Summary

The analogy between insulin levels and money growth gives an answer for the dominant reason people get fat and how people can lose weight:

  • When insulin is too high for too much of the time, people get fat. What makes insulin too high for too long is complex.

  • Just as it is very difficult to lower inflation without a recession, it is very difficult to lose much weight without substantial periods of time with essentially zero food consumption—"fasting". To those who treat the idea of skipping meals as too horrible to contemplate, losing weight and keeping it off can seem as difficult as taming inflation seemed in the late 1970's, before Paul Volcker demonstrated the efficacy of sharp recessions in taming inflation.

The bottom line is that chronically high insulin levels lead to obesity. Low enough insulin levels lead to weight loss.

The key question is then: "What drives insulin levels?"



Blood Glucose Levels as a Driver of Insulin Levels

One key driver of insulin levels is blood glucose levels. To avoid big problems, a high priority for our bodies is to keep blood glucose levels within fairly tight bounds. (Integrated over substantial period of time, this implies keeping blood glucose production and use in close balance.) The key regulatory mechanism for blood glucose levels is a predator-prey type relationship between insulin and blood glucose. When blood glucose rises, blood insulin levels rise, which brings blood glucose back down. When blood glucose falls, blood insulin levels fall, which brings blood glucose back up. 

The kinds of large oscillations shown in the image above only occur if an outside driving force has a relatively sudden impulse. Of things with a direct effect on blood glucose, the most common sudden impulse is eating sugar or some other quickly digestible carbohydrates. Other than sugar in all of its forms, including in fruit juice, what do I mean by easily digestible carbohydrates? First, I mean anything made with flour. Second, I mean whole foods with a high "glycemic index." (The glycemic index measures the blood glucose response to eating a particular type of food.) Starchy foods often have a high glycemic index. In the table below, think of anything above 40 as high. Thus, for example, rice, potatoes, carrots, bananas and watermelon (as well as baguettes and tortillas made from flour) have been found to raise blood glucose quickly.  

The Carb Rebound Effect: Why Eating Food and Drink with a High Glycemic Index Can Lead to Renewed Hunger a Couple of Hours Later

According to the logic of the predatory-prey relationship between blood glucose and insulin,

  • food and drink that raises blood glucose quickly enough causes a spike in insulin that

  • drives blood glucose levels down below normal,

  • leading to intense, insistent hunger a couple of hours later.

For example, I have noticed this when I eat Thai food with a bowl of white rice (and to a lesser extent when I eat Thai food with a bowl of brown rice). I am not the only one to have noticed that eating a bowl of white rice during a meal can lead to hunger an hour or two later. There is a genre of Chinese restaurant jokes about this, not all of them in good taste. (I have wondered whether those from ancestries that have been eating a great deal of rice for thousands of years have some genetic adaptation to reduce this effect for rice specifically. In The Obesity CodeJason Fung suggests that eating rice with something sour may blunt some of the carb rebound effect from the rice.)

The carb rebound effect is not limited to rice. The logic applies to anything that makes blood glucose spike relatively quickly.  

If you pay attention, it is easy to gather data on the carb rebound effect for you personally. (Indeed, if you work at noticing and remembering hunger an hour or two after you eat various things, your personal time series can soon contain as much information is in the aggregate time series of US GDP growth.) Whenever you feel hungry and can't quite figure out why, look at the label on the package of what you had eaten to see if there is added sugar you hadn't realized was there. 


Other Drivers of Insulin Levels

Blood glucose is not the only thing that drives insulin levels. For example, nonnutritive sweeteners do not directly raise blood glucose levels; but a great deal of research is being conducted on the suspicion that common artificial sweeteners raise insulin levels. (I would call them "artificial sweeteners," but many make a big point of being "natural" in some way. In addition to their direct effects on insulin, nonnutritive sweeteners may affect the microbiome of gut bacteria in a bad way.)  

One intriguing study had people rinse their mouths with a saccharin solution, then spit it out to test the idea that sweetness itself has some effect on insulin levels. So far the evidence against nonnutritive sweeteners is mixed, and one should worry about possible prejudice against nonnutritive sweeteners distorting scientific judgments about them. But if any nonnutritive sweetener raises insulin, it is by a pathway other than through a direct effect on blood glucose. I hope a great deal of research is being conducted for each specific nonnutritive sweetener, because sugar in all its forms definitely does mess with the insulin system.  

Jason Fung reads the evidence against nonnutritive sweeteners as quite damning, after combining evidence for the effects on obesity itself, disease, and on insulin. Here is a key passage from Chapter 15, "The Diet Delusion": 

Despite reducing sugar, diet sodas do not reduce the risk of obesity, metabolic syndrome, strokes or heart attacks. But why? Because it is insulin, not calories, that ultimately drives obesity and metabolic syndrome.

The important question is this: Do artificial sweeteners increase insulin levels? Sucralose raises insulin by 20 percent, despite the fact that it contains no calories and no sugar. This insulin-raising effect has also been shown for other artificial sweeteners, including the “natural” sweetener stevia. Despite having a minimal effect on blood sugars, both aspartame and stevia raised insulin levels higher even than table sugar. Artificial sweeteners that raise insulin should be expected to be harmful, not beneficial. Artificial sweeteners may decrease calories and sugar, but not insulin. Yet it is insulin that drives weight gain and diabetes.

Artificial sweeteners may also cause harm by increasing cravings. The brain may perceive an incomplete sense of reward by sensing sweetness without calories, which may then cause overcompensation and increased appetite and cravings. Functional MRI studies show that glucose activates the brain’s reward centers fully—but not sucralose. The incomplete activation could stimulate cravings for sweet food to fully activate the reward centers. In other words, you may be developing a habit of eating sweet foods, leading to overeating. Indeed, most controlled trials show that there is no reduction in caloric intake with the use of artificial sweeteners.

More generally, the effect of many different foods and beverages on insulin have been measured; the effects on insulin are not simply proportional to the effects on blood glucose, so other things are going on, too. Here is a link to an ungated food insulin index table for many types of food and drink. Let me copy out a few results that intrigued me, doubling the standard errors in the table to get 95% confidence intervals. Glucose itself is normalized to have a food insulin index of 100. I consider anything above 40 high. 


Food Insulin Indices


  • strawberry lowfat yogurt . 64 ± 12

  • skim milk 60 ± 26

  • 1% milk 34 ± 8

  • whole milk 24 ± 6

  • cream cheese 18 ± 12

  • butter 2 ± 2


Fruit and Fruit Juice

  • fresh honeydew melon 93 ± 30

  • bananas 59 ± 8

  • oranges 44 ± 4

  • orange juice 55 ± 14

  • red delicious apples 43 ± 6

  • apple juice 47 ± 4

  • avocado 4 ± 2

  • olive oil 3 ± 2


Other Beverages

  • fruit punch 76 ± 20

  • ice tea 69 ± 18

  • Coca-Cola 44 ± 6

  • beer (4.9% alcohol) 20 ± 8

  • white wine (11% alcohol) 3 ± 2



  • frozen corn 39 ± 14

  • peeled, boiled potatoes 88 ± 16

  • McDonald's French fries 57± 12


Grain and Flour

  • white rice 58± 18

  • brown rice 45± 16

  • white bread 73± 10

  • whole-meal bread 70± 18

  • fat-free blueberry muffin 69 ± 12

  • pizza 47± 8

  • brown pasta 29± 8

  • white pasta 29± 8


Meat and Eggs

  • eggs (poached) 23± 8

  • steak 37 ± 24

  • skinless roast chicken 17± 8

  • white fish 43± 26

  • hot dog 16± 6

  • bacon 9± 4

  • tuna packed in water 26± 8

  • tuna packed in oil (drained) 16± 4


Beans, Lentils and Nuts

  • canned navy beans 23± 8

  • lentils 42 ± 18

  • salted, roasted peanuts 15 ± 4

  • walnuts 5± 1

standard errors doubled to indicate 95% confidence intervals

Let me make several points based on this table.

  1. Although the 95% confidence interval is wide, these results back up the idea that skim milk is fattening that I hint at in my post "Whole Milk Is Healthy; Skim Milk Less So."

  2. Fruit is not innocent if one is worried about insulin elevation and the consequent effects on obesity.

  3. Fruit juice elevates insulin more than whole fruit—note the comparisons between oranges and orange juice, apples and apple juice.

  4. Avocados and olive oil are great if one wants to avoid elevating insulin.

  5. I was surprised at the results for beverages—ice tea being so high, beer and wine being so low, in generating insulin; Coca-Cola not being worse.

  6. Oil tends to reduce the effects of a food on insulin. For this, look at the comparison between boiled potatoes and McDonald's french fries, the comparison between tuna packed in water and tuna packed in oil, and the fact that pizza isn't worse.

  7. You should worry about anything labeled "fat-free," "low-fat" or "reduced-fat." Besides skim milk, the lowfat yogurt and the fat-free blueberry muffin are things to stay away from. The entire table (too big to reproduce in its entirety) reinforces this point.

  8. Bread and rice look bad, backing up what I said above.

  9. I was surprised that pasta was as low as it is in its tendency to generate an insulin response. I have no good explanation for this.

  10. Eggs look good.

  11. In general, meat is OK in terms of its effect on insulin. I was surprised that bacon and hot dogs were so low in generating an insulin response and white fish so high.

  12. Nuts look good.


Insulin Resistance and the Side Effects of Chronically Elevated Insulin

A hormone can be thought of as a drug the body itself produces. In that sense insulin is a powerful drug. Like other drugs, high doses can lead to drug resistance. Insulin resistance is a very serious problem. Indeed, insulin resistance is very closely linked to metabolic syndrome and prediabetes

Because keeping blood sugar within a fairly tight range is such a high priority for the body, when target cells become resistant to insulin, to get the job done, insulin levels simply go up more in response to any given increase in blood sugar. That is, the insulin system starts shouting if some of the cells insulin talks to become hard of hearing. 

If every effect of insulin were blunted to exactly the same degree by insulin resistance, this wouldn't be a big problem. But typically, many dangerous side effects of high insulin remain at close to full strength even when the blood-glucose-lowering effect of insulin is blunted.

When people respond to the carb rebound effect by eating more carbs when they get hungry again, easily digestible carbohydrates can lead to a chronic elevation of insulin which is enhanced in a slow-moving multiplier effect by insulin resistance. Gary Taubes, in his history of nutritional thought Good Calories, Bad Calorieshypothesizes that refined carbohydrates, with the chronic elevation of insulin they cause, could be behind many of the diseases that distinguish the bulk of those living in a modern way from those living in traditional ways that have now mostly vanished from the Earth. 

[Update, February 6, 2018: As I detail in "The Case Against the Case Against Sugar: Seth Yoder vs. Gary Taubes," I have become convinced that Gary Taubes is not a reliable reporter, so

  • I deleted the word "excellent" before the phrase "history of nutritional thought Good Calories, Bad Calories" just above, and

  • I wrote the post "Diseases of Civilization" documenting widespread agreement on the phenomenon of "Diseases of Civilization" to be explained, even though there is much disagreement about what aspects of modern diets and lifestyles causes these diseases.)

In the passage just below, Gary is in line with other sources on the description of the phenomenon of the Diseases of Civilization, and his hypothesis that sugar and other easily digestible carbohydrates play a key role in the Diseases of Civilization is a reasonable, though unproven, hypothesis.]

Here is the way Gary Taubes frames his argument in Chapter 5, "Diseases of Civilization": 

On April 16, 1913, Albert Schweitzer arrived at Lambaréné, a small village in the interior lowlands of West Africa, to establish a missionary hospital on the banks of the Ogowe River. Attended by his wife, Hélène, who had trained as a nurse, he began treating patients the very next morning. Schweitzer estimated that he saw almost two thousand patients in the first nine months, and then averaged thirty to forty a day and three operations a week for the better part of four decades. The chief complaints, at least in the beginning, were endemic diseases and infections: malaria, sleeping sickness, leprosy, elephantiasis, tropical dysentery, and scabies. 

Forty-one years after Schweitzer’s arrival, and a year and a half after he received the Nobel Peace Prize for his missionary work, Schweitzer encountered his first case of appendicitis among the African natives. Appendicitis was not the only Western disease to which the natives seemed to be resistant. “On my arrival in Gabon,” he wrote, “I was astonished to encounter no cases of cancer…. I can not, of course, say positively that there was no cancer at all, but, like other frontier doctors, I can only say that if any cases existed they must have been quite rare.” In the decades that followed, he witnessed a steady increase in cancer victims. “My observations inclined me to attribute this to the fact that the natives were living more and more after the manner of the whites.”

As Schweitzer had suggested, his experience was not uncommon for the era. In 1902, Samuel Hutton, a University of Manchester–trained physician, began treating patients at a Moravian mission in the town of Nain, on the northern coast of Labrador, or about as far from the jungles of West Africa as can be imagined, in both climate and the nature of the indigenous population. As Hutton told it, his Eskimo patients fell into two categories: There were those who lived isolated from European settlements and ate a traditional Eskimo diet. “The Eskimo is a meat eater,” he wrote, “the vegetable part of his diet is a meager one.” Then there were those Eskimos living in Nain or near other European settlers who had taken to consuming a “settler’s dietary,” consisting primarily of “tea, bread, ship’s biscuits, molasses, and salt fish or pork.” Among the former, European diseases were uncommon or remarkably rare. ...

Most of these historical observations came from colonial and missionary physicians like Schweitzer and Hutton, administering to populations prior to and coincidental with their first substantial exposure to Western foods. The new diet inevitably included carbohydrate foods that could be transported around the world without spoiling or being devoured by rodents on the way: sugar, molasses, white flour, and white rice. Then diseases of civilization, or Western diseases, would appear: obesity, diabetes mellitus, cardiovascular disease, hypertension and stroke, various forms of cancer, cavities, periodontal disease, appendicitis, peptic ulcers, diverticulitis, gallstones, hemorrhoids, varicose veins, and constipation. When any diseases of civilization appeared, all of them would eventually appear. This led investigators to propose that all these diseases had a single common cause—the consumption of easily digestible, refined carbohydrates. The hypothesis was rejected in the early 1970s, when it could not be reconciled with Keys’s hypothesis that fat was the problem, an attendant implication of which was that carbohydrates were part of the solution. But was this alternative carbohydrate hypothesis rejected because compelling evidence refuted it, or for reasons considerably less scientific?

It is not impossible to think that taking a very high dose of a powerful drug every day, year after year could cause many serious side effects. Likewise, the possibility must be considered that very high doses of insulin every day, year after year, could contribute to a wide range of diseases through its side effects on a wide range of cell types. (And of course, a few of these diseases, such as dental cavities, may result from sugar or other refined carbs in a way that does not involve insulin.)

People often talk as if obesity itself caused many chronic diseases. But other than joint problems, and the social stigma of obesity, almost all of the diseases associated with obesity could be due to the common cause of elevated insulin levels. That is,

  • chronically elevated insulin levels usually cause obesity, and

  • chronically elevated insulin levels have many dangerous side effects.

There is an interesting theoretical case in which chronically high insulin levels would be de-linked from obesity. Suppose the fat cells of someone caught in a carb rebound cycle became resistant to insulin, but his or her muscle cells retained their normal sensitivity to insulin. Because the fat cells would not respond much to the insulin signal telling them to take in glucose from the bloodstream and convert it into triglycerides and then fat, he or she would not gain much weight. But to keep blood glucose levels in line, insulin levels would have to go up even further to get the job done just from the muscle cell response to insulin. If high insulin levels cause most of the chronic diseases we associate with obesity, then while still normal in weight, he or she would be at risk for all of these chronic diseases. This is someone others might envy for being able to eat anything without gaining weight—right up until he or she died of a heart attack. 

The Implications of Insulin Resistance for Obesity

If, on the other hand, muscle cells become more resistant to insulin than fat cells, then the higher insulin levels that result from insulin resistance will lead to weight gain. Moreover, the fact that even a little food can still elevate insulin levels quite a bit will make weight loss very difficult. In saying this, I have in mind this model relating insulin levels to fat accumulation and decumulation

  • high insulin level —> accumulation of body fat

  • medium insulin levels over a substantial range —> body fat steady

  • low insulin level —> fat burning

If you have insulin resistance for muscle cells but not fat cells, even small amounts of food will prevent reaching an insulin level low enough to lead to fat burning. The only way to get insulin levels low enough to lead to fat burning may be to have a substantial period of time with no food at all—fasting.  

I should say that in addition to insulin levels, weight gain and loss are also regulated by the amount of body fat an individual already has. Presumably there are hormones fat cells send out saying "Here I am." Because of the effect of high body fat levels in restraining further fat accumulation, high insulin levels lead to a higher steady-state weight, not unending weight gain. According to this theory, current average insulin levels are the main determinant of an individuals weight setpoint or "fat thermostat." Current average insulin levels are in turn determined mainly by current diet and by the progression of insulin resistance.

Similarly, the restraining effect of low body fat levels on further fat burning means that lowering one's average insulin levels a given amount typically leads to a lower steady-state weight, not unending weight loss. In addition to the direct effect of changes in what and when one eats on average insulin levels, it may  be possible to lower insulin resistance by the equivalent of a "drug holiday": prolonged periods of very low insulin levels. The way to get prolonged periods of very low insulin levels is through fasting.


The Short-Run Blood Glucose Steady-State Theory of Fasting Done Right

The image many people have of going without food for a period of time is strongly conditioned by their experience of the carb rebound effect, with its abnormally low blood sugar levels. Absent the carb rebound effect, going without food for a period of time is much easier. 

Suppose one's last meal before beginning the fast is comprised entirely of food with a very low food insulin index. Then there will be no rebound. Instead of blood sugar dropping so low it generates intense hunger, blood glucose and insulin levels will drop just low enough that the rate at which glycogen and body fat stores are raided is just enough to replace the glucose that is used up to provide energy. That is, one will go into a short-run steady-state with a slightly low—but not severely low—level of blood glucose.  

Based on my own experience and that of a couple of other people I know who are following Jason Fung's recommendations, I can describe the phenomenology of this short-run steady-state with a mildly low level of blood sugar as one of gentle hunger that I can readily distract myself from by something as simple as work or TV. (But if I have occasion to think about food, this gentle hunger feels a bit stronger.) 


Is Fasting Dangerous?

One danger of fasting is the danger of getting one's electrolytes out of balance. This danger is straightforward to address. One can either drink bone broth during one's fast as Jason Fung recommends, or take the simpler course of taking electrolyte pills such as these SaltStick capsules I have used.

Link to Amazon page for SaltStick caps. Note: Amazon may not have the cheapest prices.

(Some other electrolyte products are loaded with nonnutritive sweeteners.)

A much bigger danger of fasting is that after a while it becomes so easy that one can get tempted to try very long periods of time lasting weeks. Sparse anecdotal data discussed in this useful PaleoLeap blog post suggests some dangers from fasting for weeks, even if it seems easy and fat stores remain. But (in many cases for religious reasons) people have fasted for one or two days frequently enough it is unlikely fasting done right for these length of times is dangerous for someone in good health to begin with. Another argument that fasting for up to 48 hours should be quite safe is likelihood that our ancestors in the Pleistocene almost surely had to weather frequent periods of foodlessness for at least that long. So we should be fairly well adapted to intermittent fasting for such relatively short periods. 

[Update November 8, 2017. There is a debate about fasting works as well for women as for men. For example, claiming that it often doesn't, see the article "Intermittent Fasting for women: Important information you need to know" by Helen Kollius. Even on this negative view, careful, moderate experimentation with what works for you should be OK for most women to see if you are one of the women who doesn't have these issues, but be alert to the warning signs in Helen Kollius's article, and don't go too fast, despite the enthusiasm you may feel from having real success at weight loss for the first time. On the other side of this debate are these passages from Jason Fung and Jimmy Moore in The Complete Guide to Fasting, Chapter 10 "Who Should Not Fast":

They write:

Virtually all studies on fasting confirm that both men and women benefit from fasting. Further, there is no particular difference in efficacy between the two sexes. ...

While women can, of course, have problems during fasting, men often have the same troubles, too. Interestingly, I’ve seen the highest success rates with husbands and wives who try fasting together: the mutual support is a big help and makes fasting far easier. ...

Fasting has been part of human culture for at least two thousand years. Are Muslim women exempt from fasting? Are Buddhist women exempt from fasting? Are Catholic women exempt from fasting? Nope! With thousands of years of practical fasting experience, none of these religions make a distinction between adult men and women, except when women are pregnant or breastfeeding.

... malnourished women should not fast, since excessively low body fat can lead to amenorrhea (loss of menstrual cycles) and difficulty conceiving. But normal-weight women do not see a significant difference in their sex hormone profile during fasting.

End of November 8, 2017 Update.]

[Update January 13, 2018. For another skeptical take on fasting for women, see Mark Sisson's post here.]

Legal Notice: Fast for longer than 48 hours at your own risk. Indeed, let me copy and say on my own behalf the disclaimer in that PaleoLeap blog post:

Since this article touches on a subject that could potentially have negative health outcomes, it’s in order to reiterate the website’s health disclaimer: All material on this website is provided for your information only and may not be construed as medical advice or instruction. No action or inaction should be taken based solely on the contents of this information; instead, readers should consult appropriate health professionals on any matter relating to their health and well-being.

Be sure to replenish your electrolytes even during shorter fasts. And sit or lie down if you ever feel faint. 

Since glycogen stores add up to only about 24 hours worth of calories, 42-48 hour fasts should be long enough to occasion quite a bit of fat burning. Even 18-hour fasts—for example, skipping lunch and snacks; eating lunch and dinner close together—are likely to lead to substantial fat burning when one's initial long-run steady-state level of body fat is based on never "fasting" longer than the 8 hours or less of sleep.  

For those tempted by longer fasts than 48 hours, a safer approach (given how little solid evidence we have about the safety of longer fasts) is to do a long modified fast broken up by a meal with a very low food insulin index every 48 hours. Then when the long modified fast is over, feast on a wider variety of foods with somewhat higher food insulin indices (but still below 40) so that the overall program retains enough food fun to be sustainable. 


But What About Calories-In/Calories Out? 

My views on calories in/calories out are well expressed in my Storify story "How the Calories In/Calories Out Theory Obscures the Endogeneity of Calories In and Out to Subjective Hunger and Energy." Including in weight gain in calories all three of glycogen (at about 450 calories per pound of glycogen plus associated water), body fat (at about 3500 calories per pound) and muscle (at about 2500 calories per pound) there is indeed an energy balance identity

Weight Gain in Calories = Calories Consumed - Calories Expended.

However, the terms on the right-hand side of this equation are not exogenous.

  1. Calories expended depend on blood glucose levels both through changes in metabolic rates and subconscious changes in activity levels.

  2. Insulin initially raises calories expended, but then drives blood glucose levels down, leading not only to lower calories expended but also to greater hunger and thus often to more calories consumed.

On the endogeneity of calories in and calories out, also see "Julia Belluz and Javier Zarracina: Why You'll Be Disappointed If You Are Exercising to Lose Weight, Explained with 60+ Studies." 

In many ways, it is more instructive to read the energy balance identity in this direction:

Calories Expended = Calories Consumed + Weight Loss in Calories

That is, for a given amount of calories consumed, the more effective one is at weight loss, the more energetic one will feel. If insulin levels are high enough to prevent weight loss, one will feel less energetic for a given amount of calories consumed than if insulin levels are low enough to allow weight loss. What this means in practice is that eating just a little bit that is mostly easily digestible carbs feels awful, while if one eats nothing one can effectively feast on one's own body fat.

Both of these experiments have been done. Calorie restriction with easily digestible carbs in the mix is very unpleasant. (More on that in a future post.) Fasting is not that bad, done right.  

During fasting, the energy balance identity becomes

Calories Expended = Weight Loss in Calories

Weight loss in calories in turn goes up with very low insulin levels. Fortunately, even insulin resistance doesn't prevent insulin levels from being very low by the end of a substantial fast. But insulin resistance does mean that the initial part of the fast before insulin levels have fallen all the way might be somewhat harder, both in less weight being lost and in how energetic one feels. 

There is a lot more to be said to trash the way the energy balance identity is misinterpreted. The main mistake is adding in false exogeneity assumptions under the table. I hope to write more posts about that. In "How the Calories In/Calories Out Theory Obscures the Endogeneity of Calories In and Out to Subjective Hunger and Energy" I make the analogy that the way calories in/calories out is usually interpreted—"To lose weight, eat less, exercise more"—is about as insightful as saying that the way to improve the trade balance is to increase exports and reduce imports. To dig into that analogy, the way calories in/calories out is usually interpreted is like taking the identity

Trade Balance = Exports - Imports

and assuming that exports and imports are exogenous except for the effects of tariffs. In reality, it is much closer to the truth to say that the adjustment of the exchange rate makes exports and imports adjust to equal desired capital flows. (See "International Finance: A Primer.") My central analogy in this post is to compare insulin levels to the money growth rate. But a secondary analogy here at the bottom of this post, the new analogy is to compare insulin levels to the international capital flows that drive trade balances. Everything else adjusts to be in line with the insulin signals.  



It is common for family practice doctors to tell their patients they should lose weight. The typical advice is "eat less, exercise more." Think of this advice "eat less, exercise more" as an intervention. If being told the words "eat less, exercise more" were a pill, we would judge this pill a public health failure. People all around the world keep getting fatter, despite being well dosed with this "eat less, exercise more" advice. 

The problem is that the advice "eat less, exercise more" treats food intake and energy expenditure as if it were an exogenous conscious decision. My view is that effective weight loss is all about understanding the body's regulatory mechanisms. (Of course, exercise has many other benefits, however little it contributes to weight loss.) 

One nice demonstration of the importance of the body's regulatory mechanism for weight is that if one gains a pound a year of fat, as is the trend throughout adulthood for many Americans, that excess of 3500 calories a year of calories consumed over calories expended is a little less than 10 calories a day. Getting things in balance within an average of 10 calories a day is quite a feat! Clearly a lot is going on at a subconscious level. To an excellent approximation, it is all about the regulatory system. 

Insulin and its countersignal glucagon are the central hormones in the human body's overall energy regulation system. Pay attention to how their levels are determined, and devising effective weight control strategies becomes much easier. 


Don’t miss my other posts on diet and health:

I. The Basics

II. Sugar as a Slow Poison

III. Anti-Cancer Eating

IV. Eating Tips

V. Calories In/Calories Out

VI. Other Health Issues

VII. Wonkish

VIII. Debates about Particular Foods and about Exercise

IX. Gary Taubes

X. Twitter Discussions

XI. On My Interest in Diet and Health

See the last section of "Five Books That Have Changed My Life" and the podcast "Miles Kimball Explains to Tracy Alloway and Joe Weisenthal Why Losing Weight Is Like Defeating Inflation." If you want to know how I got interested in diet and health and fighting obesity and a little more about my own experience with weight gain and weight loss, see “Diana Kimball: Listening Creates Possibilities and my post "A Barycentric Autobiography. I defend the ability of economists like me to make a contribution to understanding diet and health in “On the Epistemology of Diet and Health: Miles Refuses to `Stay in His Lane’.”