Calories Don't Count
"According to Democritus, truth lies at the bottom of a well, the water of which serves as a mirror in which objects may be reflected. I have heard, however, that some philosophers, in seeking for truth, to pay homage to her, have seen their own image and adored it instead."
-- Jean Paul Richter
"La fixité du milieu intérieur est la condition d'une vie libre et indépendante."
-- Claude Bernard, "Leçons sur les phénomènes de la vie" (1878)
I'm slowly learning to take a perverse delight in discovering my errors. In connection with my post on sugar and insulin, I've realized that this old post I wrote about obesity is wrong -- and not just superficially wrong, but deeply and fundamentally wrong. It ignores basic principles about how bodies work, and fallaciously presses the First Law of thermodynamics into the service of a morality play.
The fallacy is to assume that a change in energy stored in a system necessarily has to be caused by a change in energy in and/or energy out. But there is no basis for assuming this in biology: living organisms are endowed with complex machinery for maintaining a consistent internal environment, precisely because they can't afford to be at the mercy of small changes in inputs and outputs. You start using drugs to douse your brain in excess dopamine and it responds with receptor downregulation; your blood sugar gets low and your pancreas starts pumping out glucagon to raise it; your muscles start using more oxygen and dumping too much carbon dioxide into your blood stream, so your heart rate goes up and you start breathing harder to clear it faster. And so on and on -- these systems are ubiquitous and arguably part of the definition of life. So why assume the caloric inputs and outputs are drivers of change in energy storage rather than the other way around? The First Law has no causal arrow.
Having deposed one hypothesis from its privileged place and cast aspersions on its claims to the throne of legitimacy, we can repair to what we know about insulin's role in regulating fat storage for an alternative hypothesis: calories are not all created equal, the biochemical properties of what we eat have an effect on endocrine response, and endocrine response is what controls the energy economy of your body. Hormones govern when and where fat accumulates, as demonstrated by the different accumulation profiles of men and women; they also govern energy intake (and output to a significant extent), as demonstrated by the ravenous appetites of rapidly growing children. Insulin, being the only hormone in your body that shifts the savings/spending energy balance in your body toward savings, is the natural culprit whose levels we might expect to correlate most tightly with percentage body fat. It seems obvious in retrospect, yet much of what we've been told about diet for the past three decades has ignored it. (Prior to the 1970s, the conventional wisdom for well over a century was, correctly as it turns out, that starches rather than fats drive obesity.)
There's an enzyme called lipoprotein lipase (LPL) whose primary function is to regulate the uptake of fats into a cell. LPL is affected differently by different hormones, which is why women and men tend to accumulate fat in different areas. Insulin signals to your fat storage cells to up-regulate LPL, causing them to suck in more fat for storage. On the other hand, epinephrine signals to your muscles to suck in fat to be used for fuel. So even if calories in doesn't matter per se, burning calories during exercise should still be a good way to lose weight, right?
Well, no. After the exercise stops and the epinephrine goes away, the LPL balance shifts back to fat cells which proceed to reabsorb much of the fat left in circulation. Combining this with all the glucose and fat your muscles burned during exercise, you get hungry again and tend to eat to make up for what you lost. Statistically speaking, the effect is very close to being a wash: in a systematic meta-review of studies exploring the exercise-weight relationship, Fogelholm & Kukkonen-Harjula (2000) found that while exercise slowed the rate of weight regain in people who'd successfully lost weight by dieting, everyone nonetheless gained it back whether they exercised or not. And the difference in rate was itself pretty puny: 0.28 vs 0.33 kg/month.
If strenuous exercise causes appetite rebound, what about constant gentle exercise like I suggested in my old post? Turns out I should have done the arithmetic, for alas I was talking nonsense. Take the example of a 200lb (~90kg) person going up a flight of stairs 3m high. Most of the work done in this situation will be against gravity, with the horizontal dimension being negligible. The work done if muscles were perfectly efficient would be:
Work = mass * acceleration * distance = 90kg * 9.8 m/s^2 * 3m = 2646 Joules
But human muscle is only about 25% efficient, so you have to apply a corrective of about 4x, giving you 10,584J. There are about 4.2 joules in a calorie, so this would burn about 2500 calories. Not bad, right? Well, yes actually: dietary "calories" quoted on food items are actually kilocalories, so what you're actually burning is 2.5 "calories" per flight of stairs. Two-point-five. As in, one sixth the amount contained in a level teaspoon of sugar. You'd have to hike up all 108 floors of the Sears Tower just to work off a reuben. (Walking an equivalent distance horizontally burns much less since you're not fighting gravity so much. You'd have to run at least three miles to burn the equivalent amount of energy.) Non-intensive exercise has so little effect on weight compared to diet that you couldn't distinguish its effects from a rounding error. (Newburgh (1944) ran these numbers and came up with a similar result.)
* * * * *
I owe the above revelations (and more to come) to Gary Taubes, whose interviews and articles I've spent some time googling and reading, and whose book on this subject is now on my wishlist. The talk embedded below is over an hour long, but worth it IMO. Just try to ignore the fact that he uses the phrase "shift your paradigm" at the outset -- stick with it and you'll probably understand obesity better than you do now. (Everyone already hip to all this jive is exempted, of course.)
If the major angle Taubes is pushing is right (and I think the weight of evidence supports it), everyone responsible for the promulgation of the low-fat-better-health dogma should, at minimum, be publicly disgraced for committing what amounts to medical malpractice on a massive scale. Fat chance, of course: the junking of the atrocious food pyramid three years ago is a tacit admission that Taubes and the researchers his work is based on are more or less right, but I am unaware of any heads rolling.
(Those craving more info are advised to read Taubes' infamous NYT article from 2003 against the anti-fat dogma, and a more recent NY Magazine article on why exercise doesn't make you leaner.)
-- Jean Paul Richter
"La fixité du milieu intérieur est la condition d'une vie libre et indépendante."
-- Claude Bernard, "Leçons sur les phénomènes de la vie" (1878)
I'm slowly learning to take a perverse delight in discovering my errors. In connection with my post on sugar and insulin, I've realized that this old post I wrote about obesity is wrong -- and not just superficially wrong, but deeply and fundamentally wrong. It ignores basic principles about how bodies work, and fallaciously presses the First Law of thermodynamics into the service of a morality play.
The fallacy is to assume that a change in energy stored in a system necessarily has to be caused by a change in energy in and/or energy out. But there is no basis for assuming this in biology: living organisms are endowed with complex machinery for maintaining a consistent internal environment, precisely because they can't afford to be at the mercy of small changes in inputs and outputs. You start using drugs to douse your brain in excess dopamine and it responds with receptor downregulation; your blood sugar gets low and your pancreas starts pumping out glucagon to raise it; your muscles start using more oxygen and dumping too much carbon dioxide into your blood stream, so your heart rate goes up and you start breathing harder to clear it faster. And so on and on -- these systems are ubiquitous and arguably part of the definition of life. So why assume the caloric inputs and outputs are drivers of change in energy storage rather than the other way around? The First Law has no causal arrow.
Having deposed one hypothesis from its privileged place and cast aspersions on its claims to the throne of legitimacy, we can repair to what we know about insulin's role in regulating fat storage for an alternative hypothesis: calories are not all created equal, the biochemical properties of what we eat have an effect on endocrine response, and endocrine response is what controls the energy economy of your body. Hormones govern when and where fat accumulates, as demonstrated by the different accumulation profiles of men and women; they also govern energy intake (and output to a significant extent), as demonstrated by the ravenous appetites of rapidly growing children. Insulin, being the only hormone in your body that shifts the savings/spending energy balance in your body toward savings, is the natural culprit whose levels we might expect to correlate most tightly with percentage body fat. It seems obvious in retrospect, yet much of what we've been told about diet for the past three decades has ignored it. (Prior to the 1970s, the conventional wisdom for well over a century was, correctly as it turns out, that starches rather than fats drive obesity.)
There's an enzyme called lipoprotein lipase (LPL) whose primary function is to regulate the uptake of fats into a cell. LPL is affected differently by different hormones, which is why women and men tend to accumulate fat in different areas. Insulin signals to your fat storage cells to up-regulate LPL, causing them to suck in more fat for storage. On the other hand, epinephrine signals to your muscles to suck in fat to be used for fuel. So even if calories in doesn't matter per se, burning calories during exercise should still be a good way to lose weight, right?
Well, no. After the exercise stops and the epinephrine goes away, the LPL balance shifts back to fat cells which proceed to reabsorb much of the fat left in circulation. Combining this with all the glucose and fat your muscles burned during exercise, you get hungry again and tend to eat to make up for what you lost. Statistically speaking, the effect is very close to being a wash: in a systematic meta-review of studies exploring the exercise-weight relationship, Fogelholm & Kukkonen-Harjula (2000) found that while exercise slowed the rate of weight regain in people who'd successfully lost weight by dieting, everyone nonetheless gained it back whether they exercised or not. And the difference in rate was itself pretty puny: 0.28 vs 0.33 kg/month.
If strenuous exercise causes appetite rebound, what about constant gentle exercise like I suggested in my old post? Turns out I should have done the arithmetic, for alas I was talking nonsense. Take the example of a 200lb (~90kg) person going up a flight of stairs 3m high. Most of the work done in this situation will be against gravity, with the horizontal dimension being negligible. The work done if muscles were perfectly efficient would be:
Work = mass * acceleration * distance = 90kg * 9.8 m/s^2 * 3m = 2646 Joules
But human muscle is only about 25% efficient, so you have to apply a corrective of about 4x, giving you 10,584J. There are about 4.2 joules in a calorie, so this would burn about 2500 calories. Not bad, right? Well, yes actually: dietary "calories" quoted on food items are actually kilocalories, so what you're actually burning is 2.5 "calories" per flight of stairs. Two-point-five. As in, one sixth the amount contained in a level teaspoon of sugar. You'd have to hike up all 108 floors of the Sears Tower just to work off a reuben. (Walking an equivalent distance horizontally burns much less since you're not fighting gravity so much. You'd have to run at least three miles to burn the equivalent amount of energy.) Non-intensive exercise has so little effect on weight compared to diet that you couldn't distinguish its effects from a rounding error. (Newburgh (1944) ran these numbers and came up with a similar result.)
* * * * *
I owe the above revelations (and more to come) to Gary Taubes, whose interviews and articles I've spent some time googling and reading, and whose book on this subject is now on my wishlist. The talk embedded below is over an hour long, but worth it IMO. Just try to ignore the fact that he uses the phrase "shift your paradigm" at the outset -- stick with it and you'll probably understand obesity better than you do now. (Everyone already hip to all this jive is exempted, of course.)
If the major angle Taubes is pushing is right (and I think the weight of evidence supports it), everyone responsible for the promulgation of the low-fat-better-health dogma should, at minimum, be publicly disgraced for committing what amounts to medical malpractice on a massive scale. Fat chance, of course: the junking of the atrocious food pyramid three years ago is a tacit admission that Taubes and the researchers his work is based on are more or less right, but I am unaware of any heads rolling.
(Those craving more info are advised to read Taubes' infamous NYT article from 2003 against the anti-fat dogma, and a more recent NY Magazine article on why exercise doesn't make you leaner.)