Letter From The Editor - Issue 69 - June 2019

Bookmark and Share

About IGMS / Staff
Write to Us

  Science Fact-ion by Randall Hayes
November 2017

Hacking the Weight-O-Stat (with Sprinkles!)

“Shangri-La? Odd name for a diet. Name of a spa, maybe.”
--Seth Roberts, The Shangri-La Diet

A few months ago I wrote about the interactions between classical conditioning and the immune system, treating the immune response as an insanely complicated reflex. I implied, based on a small amount of research, that drug responses could be imagined the same way, and encouraged SF writers to explore the idea in story form. This month, I want to push that envelope even further, to include weight control, as proposed in an unpublished review paper and a best-selling diet book.

SF has spent a long time predicting that humans of the future might evolve away from their prehistoric muscular physicality, either by diverting ever more of their metabolic resources to their brains or, as in Wall-E, by bloating up into immobile couch potatoes who rely on technology to run their lives. I've never seen anyone write about how that might actually happen, mechanistically. Our current pop-culture model of weight control is some sloppy intuitive combination of a thermostat and a moral failure.

First, the thermostat model: when weight goes down, we get hungry, and we eat. What happens when weight goes up? A good thermostat would turn off the heat, and maybe turn on the AC. It is obvious from common experience that obese people still get hungry, but we haven't updated our mental model to reflect this. Medicine, victim to the same naïve model, has searched and searched for faulty metabolic wiring in the weight-o-stat, and there are a few rare syndromes, like leptin deficiency, that have excited hopes of an easy pill cure, but none of these can account for the current obesity epidemic. New comparative genetic studies may change this; it seems that the “thrifty gene” that was hypothesized to underlie obesity may be older than we thought. Still, genes will only be part of the answer.

Second, our primate social circuitry kicks in to assign blame, like it does with any other addiction. Maybe the weight-o-stat is overridden by culture, by advertising, or by a vague weakness of character. We may even assume that obese people are lying about their hunger, or (invoking depth psychology) that they are lying to themselves—hallucinating their hunger, in a sense.

I don't deny that there is a psychological/behavioral dimension to obesity, that we eat to comfort ourselves, or that we eat because of social demands (though just asking people about these relationships does not appear to work very well). Any behavior as biologically important as eating gets tangled up with individual learning and all kinds of cultural dynamics (which require sophisticated measures to untangle). It's precisely the multidimensional nature of the problem that I find so interesting.

Psychologist Seth Roberts, who I mentioned in a previous column on self-experimentation, stepped back to re-imagine the problem of overeating. What if, for the sake of argument, we drop our cheater-detection game theory assumptions and imagine that obese people are telling the truth about their experience of real hunger? What if the weight-o-stat is not broken? What if the weight-o-stat is working perfectly for the hungry environment in which it evolved?

We know that many mammals, such as rodents and bears and even one primate (the mouse lemur), gain lots of weight, specifically storing energy as fat, and then lose it again during hibernation or torpor. Roberts proposed that as mammals, humans might have a similar bias towards weight-gain, but controlled by different signals. We evolved near the equator, so the day-length cues that most hibernating mammals use to prepare for winter would not have worked (though those signals are so old, evolutionarily, they might still be accidentally relevant now that we have spread around the world). The fact that humans lived in famine-prone tropical environments, variable in their rainfall and such, means our ancestors needed a system that would prime them to gorge whenever food was available, on the assumption that soon it would not be. They rarely had a chance to stay fat enough, for long enough, to develop the health problems we see so often today.

Roberts further proposed that humans, as nomadic omnivores with extremely varied diets, needed a weight-gain signaling system that was flexible enough to accommodate many different foods, some of which the species had never encountered before. Every time we invaded a new continent, we found new ecosystems, full of unique foods. This is where the link to the other previous column on conditioning becomes important. We have to learn to like new flavors (anyone who honestly remembers his or her first beer will know what I mean). The more often we experience a particular flavor-calorie association, the stronger it gets, and the more fattening a particular food becomes. This leads to a number of surprising conclusions.

  1. Industrial food, which is extremely consistent by design (that being their definition of quality control), should be more fattening than more variable home-cooked food, even when made with the same calorie counts.

  2. Separating flavors and calories in time should weaken the association, reducing hunger and allowing weight loss. This was the basis for The Shangri-La Diet, which basically recommends that we eat unflavored calories between meals to reduce hunger. This scheme requires that sweet and fat be considered calorie signals, not flavors in the technical sense that the model uses the word.

  3. Randomizing the flavor-calorie associations should also weaken them enough to turn off the weight-gain signal. This was my personal favorite because it led to a wacky series of experiments where subjects lost weight by putting spearmint sprinkles on their mashed potatoes one meal, and sprinkles of another flavor, like chocolate, the next meal. Really neat SF-type thoughts there.

There might be lots of other clever ways to hack the system, ways that require only creativity and persistence. We might not have to engineer artificial hormones, or undigestible pleasure foods, or alien temporal fistulas that teleport food out of your stomach before your small intestine can absorb it (though I guess those would be cool, too).

Now, what the weight-gain signal actually was, Roberts never got specific about. He seemed satisfied with the circumstantial evidence cited in his review paper. My own hypothesis is that it’s due to a conditioned insulin release. We know that insulin can drop blood sugar and trigger hunger. I have personally experienced something like this during those periods in my life when I've had regular daily meal times.


But twenty minutes later, not so much. My longest personal fast, during an academic summer on a college campus where my parents weren't watching, lasted about three days, and hunger came and went in somewhat predictable mealtime cycles. That makes no sense in a weight-o-stat model whose only input is a measure of bodily energy reserves. Hunger should grow continuously until cut off by an input of calories, or saturate at a steady maximum level. However, short-term spikes of intense hunger make perfect sense in a classical conditioning model, where the unconditioned response is calorie-driven hunger, the conditioned stimulus is time of day, and the conditioned response is a pulse of insulin, producing a time-driven hunger. This conditioned hunger could presumably happen even in an obese person.

We've now seen three examples of conditionable behaviors, two unconscious (immune suppression and drug response) and one more or less conscious (eating). The novel/movie A Clockwork Orange famously used aversive conditioning, similar to what is used in real life with sheep-killing predators, to affect sex and violence in its human protagonist. What other non-creepy possibilities are there?

Randall Hayes, Ph.D., drives home to Kentucky every year for Thanksgiving. His mom makes the best damn cornbread on this planet or any other.




Why was this paper never published? That is a very good question, and one I wish I had the answer to.




To summarize, Slate disagrees.




Questionnaires are always fraught measures, and should always be seen as only a first step towards finding real answers.


Ecological Momentary Assessments are sort of like research tweets. Subjects are pinged at random times to report whatever the researchers are interested in. It’s about the most unbiased method anyone has come up with. If you’ve never read a systematic review, this shows the rules for deciding which studies count and which don’t, in convenient flow-chart form.



OK, maybe more than one.


“Survival of the fattest?” Another reason for a credits system in science—knowing who to blame for the bad puns. Seriously, though, folks, check this one out. Mutant apes in Europe during the Miocene, frogs storing fat in their feet, and many other wonderful examples. Imagine if your feet weighed 20 pounds apiece.


Check out Figure 7 for a schematic diagram of the process.


These rodents can be triggered to gain weight just by leaving the lights on. A lot of humans leave the lights on, too. Hmmm . . .


Siberian hamsters are also a model system for studying the effects of day length on body weight. More detailed wiring diagrams of the hypothalamus in this one.


I read this book on vacation, and it is fantastic.


This study is about people who already like beer. I predict they would get different results from a first-beer situation.


Two words: “anal leakage.”


Paired with odors of rosewood and peppermint, in this case, but a clock might work, too.


Read more by Randall Hayes

Home | About IGMS
        Copyright © 2023 Hatrack River Enterprises   Web Site Hosted and Designed by WebBoulevard.com