Tuesday, July 22, 2008

Energy Conservation: It's Not Just a Good Idea, It's the Law

Michael Eades latest blog points us to a review of Gary Taubes' Good Calories, Bad Calories by Dr. George Bray. Gary Taubes was given the opportunity to respond, and as usual, pretty much brings the wood, from the standpoint of logical clarity and consistency. I haven't read Bray's review in detail, but skimming over it I have to wonder how carefully he read the book. Indeed, he seems to essentially agree with Taubes that fat storage is driven by hormonal factors as part of overall metabolic regulation, and gives some examples where obesity results from failures in these regulatory mechanisms (which we'll also explore in subsequent posts in the Energy Regulation series). Despite this apparent agreement, Bray spends about 13 pages simultaneously trying to disagree with Taubes. Smells like cognitive dissonance, at least from my cursory reading. Taubes reply does a nice job at cutting through the fog.

Bray (like many others) seems to interpret Taubes work as somehow implying a violation or misunderstanding of the First Law of Thermodynamics, which is sort of humorous considering Taubes has a degree in physics from Harvard. Physics students pretty much get these sort of fundamental laws beaten into them from day one. In addition to Bray's review, there was a lot of noise about the First Law of Thermodynamics in response to the recently reported study about low-carbohydrate vs. low-fat diets. Being a physicist, I find misapplication of the First Law thoroughly annoying, so let's dig into this topic a bit and hopefully raise the level of understanding.

Use of the term "First Law of Thermodynamics" is a bit of historical accident. Bray actually uses the term I prefer, "Law of Conservation of Mass and Energy". Actually, "Mass" is redundant, since mass is just another word for energy, so let's shorten that to the "Law of the Conservation of Energy". The statement of energy conservation is simple: in a closed system, the total quantity of energy does not change. Energy may change "forms", e.g. the stored electrical chemical energy of battery can be converted to a light. But the total amount remains unchanged. The field of thermodynamics was largely developed in the 19th century, before we knew about atoms and such. We now understand that energy conservation in "thermodynamic systems" (consisting of very large numbers of atoms) simply follows from the more general law of energy conservation for all physical systems.

Why is energy conservation a "law"? There are many "conservation laws" in physics, and they all arise because of symmetries. Mathematically, physicists model the world via equations of motion, which basically tell how the state of the system under study changes as a function of changes in time and space coordinates. A "coordinate" is just a numerical label for a point in space (or spacetime). Suppose we're doing an experiment inside of a cubical box, 1 meter on each side. We might pick a point in the box, say the bottom left front corner, to be the "origin", labeled as (0,0,0). The top right back corner is then (1, 1, 1) in meters.

But this choice is arbitrary. I could just as easily pick any other point as the origin, say the front left corner of the parking lot, and update all of my other coordinate values accordingly. This is called a "transformation". Similarly, I could move my experiment box from it's original location. In neither case would I expect the experiment to have a different outcome. That's a symmetry: I changed one thing (coordinate origin, location of box), but it did not change the physics occurring inside the box. In this case, we would say the laws of physics are symmetric with respect to position.

A given symmetry in the equations of motion implies that some physical quantity is conserved, i.e. cannot change in a closed system. Symmetry with respect to position implies the conservation of linear momentum. Suppose I turn the box and observe the same outcome. This rotational symmetry implies conservation of angular momentum. Now let's do the experiment today, come back tomorrow, and repeat. If we get the same result, we have a time translation symmetry, which implies the conservation of energy. So basically, the "Law of Energy Conservation" arises from the observed fact that all of the fundamental equations of motion in physics are invariant under time shifts. It doesn't matter whether you look now or later, the laws governing how systems evolve in space and time are unchanged. Note that this is not the same as saying that the state of the system doesn't change, just that the laws which predict how the system goes from one state to another are not affected by the passage of time (this doesn't have to be true, it is just observed to be so in all cases so far).

Now, the above discussion is a bit watered down. The mathematically rigorous version is "Noether's Theorem", and involves differential calculus and continuous transformations. One of the best physics books I've read is Lagrangian Interaction, by Noel Doughty. Very technical, but a highly illuminating read on the power of symmetry in understanding the universe. There are many other fascinating and powerful applications of symmetry as well, one of my favorites being in Probability Theory. But we'll visit that another time.

So, to review: The First Law of Thermodynamics is just another statement of the more general Law of Energy Conservation. Energy conservation in a closed system arises because the laws of physics do not change with time. If you were to ever observe an apparent violation of energy conservation, it must be either that you are not observing a closed system (haven't taken everything into account), or you've discovered new laws of physics. The former is far more likely than the latter. For example, suppose you put some water in a cup, stuck in a thermometer, and put the whole she-bang into the freezer. The temperature would drop as time passed, indicating that the average energy of the water is decreasing. But this does not imply violation of energy conservation. Were you to also measure the net heat output from the freezer, you'd find the missing energy.

Back to our original story. Bray makes the point "Over the period of about 100 years from 1787 to 1896, the Laws of Conservation of Matter and Energy were shown to apply to human beings, just as they do to animals." That's a no-brainer given what we've learned above, since humans and animals are physical systems, ultimately governed by the same physical laws as the subatomic particles which comprise these systems. They didn't know about atoms and Noether's Theorem in the 19th century, so the explicit study of energy conservation in living organisms is understandable. But now it's not even a point of discussion, so I don't know why Bray (and so many others) keep lecturing about it. As far as anyone can tell, energy conservation is built-in to the fabric of the universe. The core issue isn't violation of this law, it's whether your metabolic theory or experiment has done a complete accounting of all energy inputs and outputs.

Energy enters the body in the form of food. In healthy people, the only way it can leave the body is through physical exertion or heat. Energy may be used in the body to fuel other biological processes ("base metabolic rate"), or it can be stored in various chemical forms. Misinterpretations seem to arise because there is an assumption that base metabolic rate and heat output are independent of caloric intake, and further independent of macronutrient composition. If you assume that intake is independent of storage and output, you can draw some strange conclusions. The body has ways of regulating total input, storage, and output in an attempt to maintain energy balance in a healthy range. As such, the output side must be related to the input side, otherwise energy regulation would be doomed to failure.

Consider a simpler example: drinking water. When we're thirsty, we drink water. The signal for thirst is generated in the brain as a function of the detected water content in the body. Too low, you get thirsty. But when you drink some water, it takes time for the water to get absorbed into the blood and signal the brain. So we tend to drink more water than we actually need; that's probably also a good evolutionary strategy, sort of "better safe than sorry". The body then has mechanisms to get rid of the excess, mostly as urine. The amount of urine we produce is clearly correlated to the amount of water we drink. If water output were independent of water input, we'd be in constant danger of either dehydration or water poisoning, depending on availability of water. Like food in Western society (and increasingly elsewhere), water is abundantly available, yet people aren't dropping dead from over-hydration because the input, usage, and output are regulated by the body. Why should we expect any different for energy regulation?

Like I said earlier, if it appears that energy conservation is violated in an experiment, such as the recent low-carb vs. low-fat diet study, the most likely explanation is that the experimenters did not measure all of the energy output. They did estimate physical activity, but it's more difficult to measure heat output. Similarly, Taubes is not saying "calories don't count", but rather that you must consider all methods of energy output when discussing energy balance. Further, you must consider the physiological mechanisms that control energy input, storage, and output, because that tells you relationships amongst them. When you do this, you find not only that output correlated with input, but also that the macronutrient composition potentially affects input, storage, and output as well. Macronutrients not only affect energy balance but other physiologically important quantities. Blood sugar, for example, is tightly regulated. If it goes too high or too low, the body has problems. So we would expect a different biological response if we eat the same calories as sugar or as fat, and of course this is exactly what is observed. It should not be surprising that high-carbohydrate or high-fat diets have very different effects on metabolism. Violation of energy conservation is not required to explain the results, just that the system has different responses to different inputs, and that the caloric content of food is only one aspect that is detected and monitored by the body.

Bray actually seems to agree with this point: "The concept of energy imbalance as the basis for understanding obesity at one level does not preclude any of the influences that affect or modify food intake or energy expenditure, including the quantity and quality of food, toxins, genes, viruses, sleeping time, breast feeding, medications, etc. They are just the processes that modify
one or other component of the energy-balance system." I think the fundamental disagreement may be whether fat storage depends sensitively on the precise balance between energy intake and output, i.e. that storage is driven by eating even a little too much. But that implies a pronounced lack of robustness in the regulatory system, one which is not observed, any more than it is in regulating water balance.

Anyway, the next time someone tells you that low-carb diets can't work because they violate the First Law of Thermodynamics, you can reply with "Low-carbohydrate diets exhibit continuous symmetry under time translation transformations, hence do not violate conservation of energy." That ought to shut 'em up.

Thursday, July 17, 2008

Even More Dissonance

A recently published study comparing various weight-loss approaches has been getting a lot of press and Internet buzz, probably because the results contradict mainstream thinking about diet and health. It's pleasantly surprising that this is getting some media coverage - usually such dissonance-inducing results are largely ignored. Regina Wilshire posted an especially amusing blog, showing how different people interpret these results. You can taste the cognitive dissonance, as each individual spins the results according to their own beliefs.

The essence of the study results is that those following a low-carbohydrate diet had greater weight-loss and improvements in blood lipids. The Mediterranean diet did well also. Both of these results are predictable from what we know about metabolic regulation, but for the mainstream, this result clearly induces significant dissonance. I particularly enjoyed Dean Ornish's attempt at reconciling this dissonance. Here's a choice quote:

I'm also very skeptical of the quality of data in this study. For example, the investigators reported that those on the "low-fat" diet consumed 200 fewer calories per day—or 10,000 fewer calories per year—than those on the Mediterranean diet, yet people lost more weight on the Mediterranean diet. That's physiologically impossible.


I think Dr. Ornish needs to bone up on his biochem. We'll hit this point later in the series on Energy Regulation, but the body very definitely has a mechanism to dump excess fat calories in the form of heat. And of course Ornish's calorie-centric focus completely ignores other regulatory effects, such as insulin's effect on fat storage. Ornish does spend plenty of time telling you all about himself, what he believes, why his particular diet flavor is superior, etc. The article reads more like an infomercial than scientific exposition. Comparison of different scientific hypotheses requires inclusion of ALL relevant evidence. Ornish heavily weighs evidence of his own creation, which (not surprisingly) supports his own preconceived notions. If you selectively weigh evidence in this way, you can come to any conclusion you want.

Here's another fun quote from Ornish: "Most people associate an Atkins diet with bacon, butter and brie, not a plant-based diet like the one I recommend." There's that "I" again. Shouldn't the diet be recommended by the evidence, not one individual? I guess Dr. Ornish is smarter than the rest of us. Maybe he would grace us with a more detailed explanation of why he's "right" given our knowledge of metabolic regulation at the molecular and cellular level?

I'm not holding my breath.

Ornish's comment also highlights one of the major origins of dissonance surrounding these recent results: the seemingly unshakable belief that saturated fat ("bacon, butter, and brie") plays a role in a wide range of disease processes. We saw in the original post on cognitive dissonance that there actually exists essentially no evidence of causality (I just confirmed this with an ex-official of the American Heart Association). For example, there may be some statistical association between saturated fat consumption and development of heart disease (particularly if you limit the observational data set), but there's no evidence at all of causality at the molecular and cellular level. Let's look at a some ways in which this association might arise:


  • Fast food is often high in saturated fat. It's also often high in total refined carbohydrates, particularly fructose. The damage wrought by increase carbohydrates (fructose is particularly good at this) and the hormonal derangement from repeated insulin spikes (and probably fructose as well) quite logically predicts an increase in heart disease. The likely high consumption of oxidized fats from deep-fried foods is the cherry on top of this sundae. Lipoprotein molecules are composed of a water soluble membrane including both proteins and fatty acids. White blood cells have a specific receptor for oxidized LDL (but not unoxidized LDL), so if your LDL includes some oxidized fat from your French fries, you should expect an increased immune response, which is known to be important in the development of atherosclerosis. So if a population has a high consumption of fast food, not only is their saturated fat consumption higher, so is the consumption of refined carbohydrates and oxidized fats. Which of these actually causes the observed increase in heart disease?
  • Grain-fed beef is known to have some nutritional issues. Grains are not the natural food of cattle, who prefer to eat leafy material, which tends to be rich in the omega-3 alpha-linolenic acid. When compared with grass-fed beef, grain-fed has a significantly higher ratio of omega-6/omega-3 fatty acids. There is a biochemical reason to believe this could increase heart disease, due to the pro-inflammatory effect of omega-6 fats. Grain-fed beef is also much higher in saturated fat, so there would be an association between saturated fat intake and increased omega-6/omega-3 ratio.
  • Grain-fed beef is also higher in total fat. Guess what - carbohydrates make cows fat too! But this fat is essentially "empty calories" in that the increased fat intake does not bring significant additional micronutrients, probably displacing calories from foods that are nutrient dense. Again, at the molecular/cellular level, there are good reasons to believe these micronutrients (like magnesium) are protective against the development of heart disease.
  • Eating a crappy diet like fast food makes people sick. Sick people tend to stay inside. If you don't go outside, in all likelihood you are deficient in Vitamin D. Vitamin D deficiency is implicated in a whole host of diseases, including heart disease. I'll bet saturated fat consumption is correlated with Vitamin D deficiency as well.


I'm sure with a little thought we could come up with several more. The point is this: associating causality with an individual statistical correlation is a very slippery slope. If you have no evidence for causality, making such an association implies that you are ignoring other possible causes WITHOUT EVIDENCE. Attempting to treat sick people based on this association could be expected to be ineffective at best, harmful at worst. And of course you wind up with the precise situation we observe today, which is that some bogus dogmatic belief blocks the advancement of science due to cognitive dissonance.

Sunday, July 13, 2008

Energy Regulation 2: Appetite

In Energy Regulation 1, I asserted that the body had many regulation mechanisms for energy intake, storage, and utilization. This regulatory network presumably evolved to maintain health over a wide range of conditions: different seasonally available foods, physical requirements, hot and cold temperatures, etc. Let's start digging in to the details of what is known, which will then should inspire some ideas on what aspects of modern life could potentially knock things askew, resulting in a situation where the body actually defends an unhealthy state like obesity.

A few caveats first. Metabolic regulation is a complex and evolving subject, and much of the knowledge is very recent (if you want to give yourself a headache, check out this spreadsheet I made trying to illustrate the various parts and their relationships). Even if you were to consider all of the available science I doubt the picture is anywhere near complete, and of course I've probably only been exposed to some smallish subset of what is known. If anybody out there finds gaps in this presentation, please fill them in via the comments. Additionally, much of the research on metabolic regulation is done on animals and extrapolated to humans. Nobody is going to do experiments where, say, they directly infuse oleic acid into the brains of people. Some of the published reviews are unfortunately vague as to whether the mechanisms discussed have been studied in humans.

One final issue is that the reviews are very focused on dietary fat, and to a lesser extent carbohydrates. But protein is essential for life, so there must be appetite and metabolic controls regulating protein intake, but this is largely not discussed. For instance, I'm guessing somewhere in the body there's something that detects amino acids and influences appetite, particularly preference for protein-rich food.

With that in mind, we'll start at the beginning. Animals eat when they're hungry. In a healthy organism, hunger is a signal that available and/or stored energy is getting low and need to be replenished. Humans have three primary energy stores: the stomach, glycogen (starch) in muscle and liver tissue, and fat (triglycerides) in adipose tissue. Now, if this system is working right, low available energy should be equivalent to low stored energy. But we're going to see it's quite plausible that conditions can arise where the body thinks available energy is low, yet excess energy is in storage.

The brain acts as a central controller, receiving various signals from the body and adjusting many different "knobs" to maintain a healthy state. Peripheral tissues also exercise some independent controls as well, e.g. the pancreas will secrete insulin in response to rising blood glucose without nervous system control. This combination of central and peripheral controls provides for both robustness and responsiveness.

The major nervous system player in metabolic regulation is the hypothalamus, an area at the base of the brain, roughly the size of an almond. The hypothalamus is the main connection between the rest of the brain and the various hormone systems of the body, sharing a private circulatory system with the pituitary gland, and projecting nerve connections to various other endocrine organs as well. The hypothalamus is also well situated to sample various chemical concentrations in the blood. Most of the brain is protected by the "blood-brain barrier" (BBB), closely-packed cells which tightly control what substances pass from the blood to the brain. But the hypothalamus is located near a region where the BBB is incomplete. It's leaky, in a sense, so the hypothalamus gets a taste of much of what's in the blood. The hypothalamus can be further divided into "nuclei", which have different sensory and control functions. Of particular interest here are the arcuate nucleus (ARC), the ventromedial nucleus (VMN), and the dorsomedial hypothalamus (DMH).

The brainstem is a close neighbor of the hypothalamus, sharing lots of neural connections. The particular region called the nucleus of the solitary tract (NTS) is the termination of the afferent fibers of the vagus nerve (afferent nerves cause signals to arrive at the brain; efferent nerves allow signals to exit the brain). The vagus nerve connects to many different organs, including those of the digestive system. The NTS appears to integrate different signals (both hormonal and nervous) and send them along to the hypothalamus. The hypothalamus does some additional integration, and projects to other brain areas involved in behaviors like finding food and eating it.

I've used the term "integrate" a couple of times. What does that mean? The neurons in the brainstem and hypothalamus receive many different signals: from other nerves, from hormones like insulin, and can directly sense nutrients like glucose. The "decision" of whether the neuron fires or expresses certain proteins must factor in all of these signals. For example, the brain requires a certain blood glucose concentration to function properly. If glucose falls, regardless of the level of insulin, the brain should take some action (like stimulating appetite), because otherwise you'll die.

The ARC in particular contains two populations of special neurons. One of these expresses cocaine- and amphetamine-related transcript (CART) along with pro-opiomelanocortin (POMC). These neurons seem to be associated with appetite suppression. For instance, POMC can be chopped up to yield alpha-melanocyte-stimulating hormone (alpha-MSH), which in turn binds to the melanocortin-4 (MC-4) receptor. Genetic problems causing defects in the MC-4 receptor result in obesity characterized by overeating. The other population expresses agouti-related protein (AgRP) and neuropeptide-Y (NPY), both of which increase appetite. If NPY is infused into rat brains, they respond with a several-fold increase in food intake that lasts 6-8 hours, similar to rats that have been fasted for 36-48 hours.

Having two opposing systems (as opposed to just one that gets turned up or down) allows for rapid fine-tuning of metabolism; this idea of opposing systems which maintain balance is found elsewhere, e.g. the sympathetic and parasympathetic endocrine systems. Both classes of neurons appear to "detect" both available energy in the blood as well as hormonal levels and probably nerve signals, with opposing results. Energy nutrients go through part of the same cycle used to actually generate energy, and the resultant metabolic products appear to trigger opening/closing of ion channels on the cell membrane. Hormones like leptin and insulin have similar effects, hence the "integration" of these signals. Changing the balance of ions inside and outside the neuron affect the "action potential", make it more or less susceptible to firing, expressing proteins, etc.

NPY neurons, for example, are glucose inhibited (GI), meaning the more glucose is around, the less active they beome. If blood sugar falls, the NPY neurons become more active, and as we saw above, NPY appears to strongly stimulate appetite. So blood sugar falls, and you get hungry. Similarly if insulin or leptin falls, these neurons are activated, and again you get hungry. But what if insulin is high AND glucose is low? Well, the brain needs a certain glucose level to operate, so I would guess that the low glucose wins, because the alternative is a hypoglycemic coma and death. Have you ever had a major blood-sugar crash a few hours after a large carbohydrate-laden meal? It's the "Chinese food makes you hungry an hour later" thing (see e.g. Teriyaki Stix Beef Bowl: 102g of carbohydrate, probably all highly refined). I would bet the extreme feelings of hunger (kind of like you were starved for 36-48 hours) is the result of increasing NPY concentrations in the hypothalamus, in turn triggered by low blood glucose, even though your insulin is still elevated. Rats show a preference for high-carbohydrate meals when stimulated with NPY. If the same is true for humans, then we shouldn't be surprised that a blood-sugar crash sends us scurrying for the vending machine to fearlessly slay and consume a candy bar, regardless of how much energy is stored in the stomach or fat. So we begin to see how the system can be broken to store excess energy, mainly fat.

The scenarios described above relate more to the instantaneous availability of energy in the blood as opposed to the amount stored. The major energy store (in terms of calories) is white adipose tissue (WAT). Fat cells, or adipocytes, are not passive buckets, but rather metabolically active both in the storage/release of fatty acids as well as the secretion of hormonal signals relating to appetite and metabolic regulation. The best-known of these is leptin, a hormone whose secretion is proportional to the amount of stored fat. Leptin suppresses appetite, probably via multiple actions. Leptin inhibits the NPY/AgRP neurons (which stimulate appetite) and excites POMC/CART neurons (which decrease appetite). Leptin also slows gastric emptying, the rate at which food leaves the stomach and enters the small intestine. So more leptin (everything else being constant) should keep the stomach fuller for a longer time, and the stomach of course sends it's own signals relating to appetite and satiety. Leptin may also increase base metabolic rate via diet-induced thermogenesis, a topic we'll explore later. There is a genetic defect that causes people to secrete little or no leptin. Individuals with this genetic problem tend to overeat considerably, and extrapolating from rats may additionally have a lower metabolic rate, with the result of extreme obesity. Administration of leptin to these individuals substantially aids this condition.

Fat cells secrete other hormones as well. Adiponectin secretion is inversely correlated with stored fat: more fat, less adiponectin, and vice versa. Adiponectin has potentially influences many things, including appetite, insulin sensitivity, inflammation, and vascular function. Interleukin-6 (IL-6) causes insulin resistance in fat cells, which tends to make them release fat instead of store fat. The hypothalamus also expresses and contains receptors for IL-6, particularly in areas controlling body composition. Fat cells also express tumor necrosis factor alpha (TNF-alpha), which inhibits lipoprotein lipase (the enzyme required to get fat out of lipoproteins and into fat cells), stimulates breakdown and release of triglycerides in fat cells, and may also induce insulin resistance.

So we see mechanisms in place to control fat storage through appetite. As more fat is stored, more leptin is secreted, which should blunt the appetite. As fat is lost, leptin levels drop, which should promote appetite. Leptin (and other hormones from fat cells) may additionally modulate metabolic rate, to encourage fat burning when there is an excess, and discourage it during a deficit. So again there's a lot of knobs to turn, all aimed at maintaining fat storage in a particular range.

Our final stop is the gastrointestinal (GI) tract along with the closely related pancreas. When we eat, food hits the stomach, which does a nominal amount of digestion both mechanical and chemical. The stomach represents short-term energy storage, more or less the "gas tank" for the body, and so it's no surprise that the stomach is involved in appetite as well. Indeed, most people think of appetite in terms of "my stomach is full/empty", but we've seen above that many other factors come in to play as well. The stomach signals the full/empty state both through nerves and hormones. Stretch receptors on the stomach wall send signals via the vagus nerve indicating fullness. The stomach also secretes the hormone ghrelin, which strongly stimulates appetite. Empty stomach means more ghrelin, full stomach means less. Increasing ghrelin increases brain concentrations of NPY. So an empty stomach definitely tends to increase your appetite, but gastric signals must be integrated with the variety of other signals to actually determine the degree of appetite stimulation.

Most of the hormonal action occurs in the small intestine and pancreas, and indeed there is some interplay between these organs. The pancreas is not only an endocrine organ (which sends hormones into the blood), it is also exocrine, emitting various substances like enzymes and bile salts require to break down food so it can be absorbed through the small intestine. The small intestine itself secretes a several hormones in various quantities, depending on the total caloric content as well as the individual levels of carbohydrate, protein (really amino acids), and fat. These hormones have a wide variety of effects, including stimulation/inhibition of pancreas endocrine and exocrine functions, modification of the rate at which food passes through the GI tract, metabolic control, and of course appetite. I'm not going to cover nearly all of these hormones or their effects. Check out the spreadsheet, or this paper and this paper for details.

A major hormone secreted by the small intestine is cholecystokinin (CCK, and no, I don't know how to pronounce it). Dietary fat and protein more potently stimulate of CCK release than does carbohydrate, and long-chain fatty acids seem to have a greater effect than short-chain. CCK affects a number of systems, e.g. inducing gallbladder contraction (to release the bile needed to digest the fat which stimulated CCK release in the first place). CCK also slows gastric emptying. Again this makes sense from a regulatory standpoint. Once the small intestine has received some energy nutrients, CCK signals the stomach to stop sending more until the present batch is done processing.

CCK also strongly suppresses appetite. In rats, administering CCK reduced food intake in a dose-dependent manner: more CCK, less food eaten. The exact mechanism is unclear, but it seems to be a combination of reduction in gastric emptying (stomach stays full) and detection by the nervous system. In both monkeys and humans, the fullness of the stomach seems to modulate the appetite suppression of CCK. The afferent fibers of the vagus nerve as well as the brainstem express CCK receptors. The Otsuka-Long-Evans-Tokushima fatty rat (try saying that 3 times fast) is a genetic variant which lacks the CCK-1 receptor, and both overeats and becomes obese.

A few notes on other GI hormones. PYY-36 is released in proportion to calories and meal composition, with fat resulting in higher concentrations than protein or carbohydrate, and may inhibit food intake. Glucagon-like peptides GLP-1 and GLP-2 are cleavage products of preproglucagon. GLP-1 increases insulin secretion and suppresses glucagon release. It also slows gastric emptying and inhibits food intake. Key areas of the brain such as the ARC express GLP-1 receptors. GLP-2 release is potently stimulated by fat and carbohydrates, and may enhance the digestive and absorptive capabilities of the small intestine. Oxyntomodulin (OXM) is released in proportion to calories ingested. It suppresses appetite and gastric motility, enhances insulin secretion, decreases food intake, and possibly increases metabolic rate.

So the takeaway here is that the GI tract sends numerous hormonal signals indicating energy is present and being absorbed, please don't send any more. One interesting side-note is that the levels of some of these hormones, notably PYY-3-36, GLP-1, and OXM, all increase after gastric bypass surgery. The effect of this should be to suppress appetite, and possibly increase metabolic rate, which would explain the success of such surgeries to reduce obesity. I find this interesting, because by itself I would guess reduction in stomach size should probably have little effect on overall food intake because of the other mechanisms regulating appetite based on stored and available energy. But diddle the relevant hormones, and voila, sustained appetite reduction and weight-loss. Hopefully the increasing understanding of these regulatory mechanisms will give rise to better treatments, since surgery seems an extreme way of accomplishing the desired effect.

Finally, we come to the pancreas. The best-known pancreatic hormone is insulin, arguably the Big Mama of metabolic regulation. It is interesting to note that the protein structures of both insulin and NPY are remarkably conserved across evolution. If you look at a primitive animal like a hagfish, it's insulin and insulin receptors are fairly similar to that of humans, so much so that hagfish insulin significantly stimulates human insulin receptors. The implication is that the role of insulin is central in metabolism and development, and fairly successful as relatively drastic changes across species required little mutation of insulin. We're most familiar with insulin's role in controlling blood sugar, both by increasing tissue uptake of glucose and by regulating glucose output from the liver. Insulin also regulates many other aspects of metabolism, like fat storage and cell division. Subsequent posts will visit these in greater detail.

Insulin is manufactured by the pancreatic beta-islet cells (B-cells). When glucose enters the B-cell, it is metabolized to ATP, the primary short-term "energy currency" of the body. But rather than using that ATP for energy, some of it closes potassium ion channels. This depolarizes the cell membrane, allowing calcium ions to enter the cell and causing stored insulin to be released. The presence of glucose in the cell additionally signals the cell to manufacture more insulin. Amino acids also trigger insulin release to varying degrees, depending on the particular flavor, as do ketone bodies. The effect of fatty acids is complex and not well understood. It appears that fatty acids are necessary for normal glucose-stimulated insulin secretion. Increasing fatty acid concentrations in the short term (1-2 hours) will cause more insulin to be released for a given glucose concentration. But long term, elevated fatty acids impair insulin secretion. Both the nervous system and other hormones also affect the amount of insulin released.

Insulin affects appetite, both directly and indirectly. The indirect path involves sensitization of the body to other satiety signals like CCK (a role shared with leptin). Insulin also appears to directly signal the hypothalamus, increasing activity of POMC/CART neurons and decreasing activity of NPY/AgRP neurons. We all know that the pancreas secretes insulin in response to blood glucose, but insulin secretion is also modulated by a number of other factors. We saw above how some GI hormones potentiate greater insulin release (the so-called incretin effect). Insulin levels are also a function of body-fat: the more fat that is stored, the higher insulin is in all states (fed, fasting, etc.) So insulin signals both energy availability and energy storage, but the primary effects indicate that over the long term insulin (along with leptin) signal how much fat is stored.

If insulin is infused directly to the brain (of a rat, presumably), the result is a decrease in food intake and loss of body weight in a dose-dependent manner. If insulin receptors are blocked, food intake and body weight increase. When insulin levels in the brain are held constant over long time periods via slow infusions, animals modify their diet and body composition until a certain body weight is achieved, and that weight is subsequently defended at a level determined by the insulin concentration.

We saw an example above where high insulin could be overridden by low blood glucose to cause hunger. Insulin suppresses appetite only when blood glucose is maintained at a proper level. Insulin-induced hypoglycemia (whether from a high-carbohydrate meal or administration of insulin) triggers an override mechanism in the brain, inducing hunger and eating to avoid going into a coma. Type 1 diabetics have the opposite problem: high blood glucose and low insulin. Type 1 diabetics are typically ravenously hungry despite high glucose, again showing the integrative capacity of the brain; yet they will fail to gain weight regardless of how much they eat, as the lack of insulin disrupts other metabolic functions.

The pancreatic B-cells also co-secrete another hormone called amylin. Insulin and amylin a secreted in a fixed molecular ratio of about 10 to 100 to one. Various disease states (including obesity) and pharmacological interventions increase the amount of amylin relative to insulin. While insulin appears to primarily signal stored fat levels, amylin signals both the amount of stored fat and energy availability from food intake. Amylin is secreted in proportion to body fat and meal size. Giving rats a does of amylin prior to a meal reduces meal size. Blocking amylin receptors produces a long-lasting increase in food intake and fat storage. Amylin appears to act in the area postrema (AP) of the hindbrain. AP neurons activated by amylin are also activated by glucose, CCK, and GLP-1, and the AP projects to the NTS, which in turn projects to areas of the hypothalamus regulating appetite and metabolism.

Type I diabetes occurs due to destruction of the pancreatic B-cells, so Type I diabetics also lack amylin, which is thought to contribute to their large appetites. Type II diabetics treated with insulin often gain more weight (duh), but this can be mitigated be treatment with an amylin analog. Some doctors are prescribing amylin analogs in obese patients who are not being treated with insulin. Since amylin serves both as a satiety an adiposity signal, this works as expected: these people both eat less and lose fat. But just as most obese people are insulin resistant, they're also amylin resistant. Administration of amylin to an already overtaxed system is, I think, a questionable long-term strategy. Additionally we know the body wants to keep the insulin/amylin ratio fixed, probably for a good reason. Adding exogenous amylin to the mix perturbs this balance even more than it already is, rather than helping to restore it to a healthy state.

Last, but not least, is glucagon, manufactured and secreted by pancreatic alpha-islet cells (A-cells). Metabolically, glucagon tends to counter the effects of insulin, e.g. increasing glucose output from the liver. Glucagon secretion is stimulated mainly by protein, possibly by fat, and not at all by carbohydrate; indeed, glucose inhibits A-cell glucagon secretion. Pancreatic hormones are dumped into the portal vein, so the liver gets first shot at them. Apparently the liver removes most of the glucagon, and rather little makes it into systemic circulation. Even so, glucagon acts as a satiety signal. Rather than acting directly in the brain, glucagon's action probably occurs in the liver, which then sends a signal to the brain via the vagus nerve. Animals whose afferent vagal nerves have been blocked do not have their feeding inhibited by glucagon.

So let's see how some different meals may affect appetite. We'll revisit these later, after we've gone through the other aspects of metabolic regulation and can look at the big picture; but the isolated effects on appetite are still interesting. These are my guesses, not proven by any scientific research. Feel free to add your own scenarios to the comments.

We discussed above what may happen after a high-carbohydrate low-fat meal, like the Teriyaki Stix Beef Bowl (102g carbohydrate, 33g protein, 7g fat). The stomach fills, reducing ghrelin secretion and sending the "full" signal to the brain. The large amount of refined carbohydrates should elicit a large insulin and amylin response, further potentiated by the release of hormones like GLP-1 and OXM. The protein in particular stimulates CCK release, which along with insulin, amylin, and other GI hormones suppress appetite. But the major insulin release induces hypoglycemia. The initial effects probably are an increase in gastric emptying via nervous system control to try and balance out the blood sugar, but of course this tends to raise insulin even more. Sooner or later the depressed blood glucose causes an increase in brain NPY, and powerful hunger, despite the fact that rather little of the meal may actually have been used for energy.

How about a "healthy" low-calorie meal, maybe a really big salad, lots of veggies and fat-free dressing. The conventional wisdom is that the large fiber load (and amount of water) fills up the stomach, contributing to satiety. That's true, to a certain extent, as filling the stomach triggers both the stretch receptors and reduces ghrelin. But the relative lack of any energy nutrients implies correspondingly low secretion of appetite control hormones like CCK. The brain also will not detect much rise in blood sugar or fatty acids. In turn, gastric emptying and intestinal motility is not inhibited and may in fact be accelerated, so the stomach empties faster than it would in a high-calorie meal. Appetite suppression from stomach distension rapidly fades, and you're hungry again.

How about a "cardiac arrest" meal of a big steak smothered in mushrooms and butter? This is a calorically dense meal, probably occupies considerably less stomach volume than the big salad, so maybe we don't get as much from stretching the stomach. But once this hits the small intestine, we get should get lots of hormones like CCK and glucagon to suppress appetite and gastric emptying. Some insulin and amylin are secreted as well. The glucagon helps keep blood sugar stable, and the additional protein from the steak may temporarily bump up blood sugar as well. The fat makes it into the circulation more slowly, and should help to both suppress appetite and gastric emptying over the longer term. Additionally, fat sensing by the hypothalamus also help the liver regulate blood sugar, so we don't get the "low glucose" override.

So to summarize:
  • The high-carbohydrate high-calorie fast-food meal makes you get hungry faster due to insulin-induced hypoglycemia. This happens despite consumption of plenty of energy.
  • The low-calorie low-glycemic salad fills you up in the short term, but you get hungry again quickly simply due to lack of available energy.
  • The high-fat high-calorie meal suppresses appetite for a longer time, both by avoiding adverse conditions like hypoglycemia, as well as providing a measured release of energy into the blood via the small intestine.
Now I ask you, which of these meals is the most likely to cause fat gain?

These examples are interesting (so I think), but again must be considered in the larger context of metabolic regulation. Obesity is not a simple result of overeating, fat-loss not the simple result of undereating. Both are a combined effect of different regulatory mechanisms. Appetite is just one piece of the puzzle. Genetically-modified rats, for example, illustrate different behavioral and physical outcomes depending on the nature of the mutation. Some overeat and maintain normal body weight. Some eat normally and get fat, and some both overeat and get fat. Conversely, to lose fat almost certainly requires restoration of the proper regulatory balance. By itself, the recommendation to "eat less and move more" is meaningless. We need to consider the effect on the hormonal and nervous system mechanisms, which require detailed thinking about the effects of food and exercise on human biochemistry.

Monday, July 7, 2008

Dissonance Redux

Michael Eades' most recent blog gives an outstanding example of cognitive dissonance in action. In short, a recent study found that obesity in China was correlated with higher intake of vegetable foods. But "everybody knows" that vegetables are healthy, right? So the authors conclude it can't be the vegetables causing obesity, it must be the vegetable oil the Chinese cook them in. The funny (but simultaneously sad) thing is that their own published data not only refutes this hypothesis, but also clearly supports the alternative hypothesis that refined carbohydrate consumption leads to hormonal imbalances which in turn lead to obesity. It's right there on the page. Talk about cognitive dissonance.

Dr. Eades did an excellent job hitting the big points, so I just want to add a few thoughts. I can't get access to the original paper (I have no interest in handing Nature any of my money for this paper, especially given that someone was clearly asleep at the wheel to let it past peer-review), but I would infer that it is an epidemiological study. So we can't give it a lot of weight, any more than we should give much weight to the mostly epidemiological evidence that support current mainstream dietary recommendations. That said, the authors' conclusion is clearly goofy, and ignores considerable information. The differences in fat consumption between less and more obese groups was fairly minimal: essentially zero in men, and about 6g in women. The statistical measure of the "trend" (indicating correlation between quantities) indicates lack of signficance in correlating fat intake with obesity for both men and women. But here's a list of items whose trend was indicated at greater than 99.9% confidence (the "+" and "-" indicate positive or negative correlation):
  • Fresh vegetables (+)
  • Fruits (+)
  • Rice (-)
  • Wheat flour (+)
  • Whole grain (+)
  • Root vegetables (+)
  • Pickled vegetables (+)
  • Fish (+)
  • Milk (+)
  • Eggs (+)
  • Calories (+, women only)
  • Protein (+)
  • Carbohydrate (+)
  • Plant food fats (+)
  • Animal food fats (-)
  • Vegetable oil (+, women only)
  • Physical activity (+)
Thus, people who ate more vegetables were more obese, people who ate more fat from animal sources were less obese, etc. Indeed, if you look at the actual amounts of food consumed, what we find between the least and most obese quartiles are major increases in wheat flour (5x men, 20x women), whole grain (18x men, 22x women), root vegetables (4x men, 4.7x women), and milk (75x men, 92x women). Physical activity in the most obese quartile was somewhat less than two times greater that in the least obese. In other words, those people following a diet and lifestyle (at least given the presented data) similar to that recommended by the USDA were the most obese, on average. The focus on vegetable consumption is actually a little weird, being only about 50% higher in the most obese quartile, compared to the least obese, considerably less variation than than that for wheat etc. I guess it's more dissonance: "everyone knows" that wheat, whole grains, etc. are healthy, so why even go there?

The ultimate scientific test of a theory is its predictive power. The theory underlying USDA recommendations basically says if you eat like the food pyramid and exercise more, you should be at lower risk of obesity. When confronted with data which contradicts the theoretical predictions, you have two choices: question the data, or question the theory. These guys did neither, instead waving their hands and adding an additional hypothesis which still failed the predictive test for one study group (men), but which I guess let them sleep at night.

Let's now consider an alternative hypothesis: that most obesity is a symptom of an underlying hormonal imbalance caused by overconsumption of refined carbohydrates. This theory predicts precisely the results seen, without any bogus ad hoc additions. Carbohydrates drive insulin drive fat storage. For that matter, milk proteins may also have a larger effect on insulin than other proteins, and of course milk does add to the carbohydrate load. There's even evidence supporting Gary Taubes' hypothesis on the connection between caloric intake and physical activity: those with higher caloric intake were, on average, more physically active. There's no evidence of causality: it may be that eating more calories increases activity, or that those with increased activity get hungrier, or both. While the weight supplied by this study is rather thin, it does at least provide further confirmation of what we would expect given current knowledge of metabolic regulation, which is considerably better than the nonsensical conclusions put forth by the authors.