In an interview with Andreas Eenfeldt (some time in 2010, uploaded by Eenfeldt 3/15/11), Gary Taubes makes the following statement (~3 minute mark is a good place to start):
Look it's 2010 we have a disorder of excess fat accumulation, and people are saying that the actual physiological regulation of fat tissue is irrelevant.
One point Stephan made in his retracted response to Gary's recent blog post needs to be addressed. So I'll do so here. Folks have been letting Taubes get away with this for a very long time, and that is his indiscriminate use of the word "regulation". The core principal of TWICHOO (Taubes Wrong Insulin-Carbohydrate Hypotheses of Obesity) is that insulin is the primary regulator of fat tissue metabolism, and thus fat mass. He's actually ratcheted that up a notch so that in his view insulin acts as virtually the sole regulator.
I'm going to use a computerized multi-fuel furnace with a fuel tank system depicted below to make my analogies.
This analogy will be a simplistic one to reflect glucose (fuel A in blue) and fatty acids (fuel B in pink). The furnace can operate on a wide range of fuel compositions but must have some minimum amount of A in the mix. A quick explanation of the furnace: On the upper right, we have the intake of A (blue 1) into a small holding tank. The fuel is either directed immediately towards the outlet tube for burning (green 1) or diverted into an overflow system (blue 2). At this point it can go to the overflow tank -- analogous to glycogen (blue 3) or to a converter (blue 4) which represents de novo lipogenesis essentially converting it to fatty acids, fuel B. Once converted to fuel B the fuel can either be added to the more immediate use pool of B or long term storage overflow B (gray 1). On the left hand side we have the intake of B (pink 1) into a small holding tank. It can either be directed immediately to the outlet tube for burning (green 2) or diverted into the overflow tank (pink 2). Note the break in this tank is intended to indicate the enormous size of this reservoir compared to the other tanks. When the furnace is in use it derives fuel from all four tanks, including directly from the intake tanks for A & B as well as the overflow tanks for each, analogous to glycogen (green 3) and adipose tissue (green 4).
Each arrow on the diagram represents a point at which there is a valve/pump all of which are at the direction of the central computer. There are also fuel gauge sensors in each tank that relay this information to the computer. In addition, there are "emergency" sensors that send priority signals if levels go below minimum or exceed maximum levels. Thus the computer in this furnace is what ultimately regulates the level of the Fuel B overflow tank -- the analogous tank to fat tissue accumulation. Overfilling this tank will result in spillage of this toxic fuel, so the computer is programmed to make sure fuel is available for the furnace to burn, but not so much as to result in toxic spillage.
We can think of each of the four green arrows as flow control valves that govern the composition and source of the fuel mix. A is burnt preferentially vs. B when in high supply, when A supply is low, more B is released and used. It is perfectly consistent with well established science to envision a single signal mediating relative flows for green 1-4, and that this regulation can be controlled almost exclusively by sensors in the A intake tank. Our analogy is insulin regulating fuel partitioning in response to carbohydrate intake. Fuel A available? Increase flow through green 1, decrease flow from green 2-4. Waiting for the next fuel A delivery? Increase flow though green 3 to maintain minimum A requirements, and green 2&4 to burn more B.
Thus insulin regulates fuel partitioning, and in the most simplistic analysis, carbohydrates by stimulating insulin ultimately regulates fuel partitioning. In our furnace analogy, we can envision that a sensor in A intake, with some feedback from other tank sensors, can do this.
This, friends, is NOT insulin regulating fat accumulation. In essence, the computer is programmed to increase/decrease flow at various points -- the computer is "passive" in this capacity, only regulating flow in response to whatever the sensors say.
Now let's look at the overall furnace operation. The furnace heats a room. It cannot run with less than, say 20% fuel A. The total amount of fuel used is determined by the temperature set on the thermostat, and the temperature of the environment outside the room. Again, for simplicity, let's envision a scenario where we are trying to keep the room temperature some degrees above the outside environment. When it's colder outside, the furnace must work more to keep temperature at a set point.
Now remember, fuel B is very toxic. As vast as the overflow tank is (let's call it OB), it has limits, and there's even some sort of spillage to a "fuel pan" as levels exceed maximum before the overage ultimately leaks into the surrounding room. Since overages of intake for both fuels ultimately result in filling up OB, this is the tank that the central computer is ultimately tasked with regulating. How would this be accomplished? Well ... we need a way for the computer to sense the level in this tank that can ultimately control the fuel balance. In other words, to keep (A+B)IN = (A+B)OUT over the long term. Because the laws of thermodynamics, that tired old tautology (grin), dictate that if (A+B)IN > (A+B)OUT the OB will fill up and overflow. By contrast, if (A+B)IN < (A+B)OUT the furnace will run dry. When you look at it this way two things become clear:
- No matter what, the central computer has no ultimate control over A&B intake. If finances are tight, you may well be forced to cut back on fuel deliveries. The furnace can be programmed to sense this and automatically reduce fuel consumption by overriding the desired temperature on the thermostat and heating the room to a few degrees lower. When prices are low over the summer, the computer can't keep you from trying to fill the tanks past capacity, though it can try to deal with your stupidity by heating the room up to a few degrees higher than what you've set on the thermostat.
- The OB level is not "regulated" by the intake of "A". It can't be in practical terms.
What would ultimately regulate the OB level? Hmmmm .... Well, we would need signals from OB to be sent directly to the central computer that controls all the valves and workings of all those arrows so that the computer could, via programming, alter various parameters so as to increase or decrease the OB level. In other words, *regulate* it. In this capacity the computer is taking an "active" role. The computer is doing more than just flipping switches inside, it is acting to alter the furnace's relationship with the environment. Either by plugging the intake ports entirely and/or increasing the fuel needs/expenditure.
If only we had an analogy in living humans that would sense the levels of fat mass (OB) and tell our brains (central computer) to set various events in motion to keep the fat mass within certain limits. Ahhh, but we do. It's called leptin. Adipose tissue is not a passive depot for fat, it is, as we all acknowledge, an endocrine organ of its own. Fat cells secrete leptin, signalling our brains as to the levels of fat in storage. Leptin does two things (there are probably more, but these two are the most relevant to the analogy and weight regulation).
- Suppresses appetite: Filled fat cells try to tell you to eat less, your body has enough. For the furnace, the OB sensor would trigger a switch in the input valve that would refuse intake.
- Increases energy expenditure: Leptin dials up REE/RMR/BMR in an effort to increase caloric expenditure to prevent fat storage from exceeding capacity. For the furnace, the OB max line sensor signals the computer to increase the thermostat setting.
So ... yes ... fat tissue does regulate its mass (or level of accumulation). It does it via leptin. Leptin acts on the hypothalamus. When leptin signaling is working properly, it tells you to eat less and it burns off minor excesses. All in an attempt to keep CI = CO and fat mass in balance. But as with all homeostatic mechanisms there are limits. We either lose our sensitivity to the signals, ignore them, or override them. A discussion way too complicated to discuss here and probably where I differ most with the more academic theories on obesity. (In a nutshell, I think we humans override this way too easily)
But what of the hyperinsulinemia associated with obesity? It could be represented by some sort of pressure valve on that OB. The last ditch compensatory action of the central computer is to try to contain the excess fuel in the OB. It just so happens that insulin is the hormone that can do this, and so there are sensors in that spillage "fuel pan" that send a signal to try to prevent the spillage despite the tank being full. These are all the same actions that control the valves to begin with. Shut down pink 2 and green 4, but the pressure causes leakage nonetheless.
There's also another hormone fat tissue generates: acylation stimulating protein, ASP. Were it not for the fact that this protein acts locally on the tissue that secretes it, it would be classed a hormone. It does much the same thing to adipocytes as insulin: e.g. enhances glucose transport into the adipocyte and stimulates esterification of fatty acids into triglycerides. Indeed it's effect on fat deposition seem to dwarf those of insulin. ASP is to intake fuel tank B of the furnace what insulin is to intake fuel tank A. It does not regulate the levels of B, but it does essentially control the pink 2 arrow. However since it is produced by adipocytes in response to dietary fat (chylomicrons) to facilitate the uptake of fatty acids by the fat tissue itself, ASP is more integral to fat tissue than insulin from the point of view of the "lipophilia" hypothesis -- that being the notion that the fat cells go rogue hoarding fuel.
Bottom line, CICO does not regulate fat mass, but fat mass does increase when intake exceeds expenditure. But neither does insulin regulate fat mass. Rather insulin mediates fuel usage for that which we've taken in to meet expenditure needs. ASP seems mostly to be responsible for keeping fat out of circulation when it presents itself in quantities requiring clearance from circulation. Ultimately it is leptin that is produced in quantities directly correlated with fat mass. Leptin is the regulating hormone that signals the hypothalamus, the regulator. The hypothalamus then signals the body to up/down regulate things any number of things (including insulin secretion) so as to increase/decrease accumulation. Leptin is how fat tissue *tries* to regulate its mass. How leptin signalling goes awry is also a subject for another time.
So Mr. Taubes. Here are my questions, with a little redundancy for effect. We are now in the latter third of 2011 discussing obesity and the physiology of the regulation of fat tissue.
How, sir, can your hypothesis and discussion the regulation of fat tissue, not include those hormones secreted by the fat tissue itself?
How can you continue to ignore leptin and ASP?
How can any hypothesis on the regulation of fat accumulation not include ASP and leptin?
Tick tick tick...