I was saddened last month to learn of the untimely death of my friend Seth Roberts, a highly original thinker on matters that lie at the intersection of psychology and physiology   Seth was known to many as a pioneer in the burgeoning field of self-experimentation. He was an early contributor to the Quantified Self movement, which owes a huge debt to his thinking. His experiments looked at how to optimize weight loss, mood, sleep, mental speed, balance and even to address specific conditions like acne. He wrote a blog on topics related to health and scientific method. I greatly enjoyed side discussions with Seth at the Ancestral Health Symposium meetings and personal correspondence over the years. This August, both Seth and I were scheduled to talk at AHS in Berkeley. While my talk on myopia will still happen, we will never hear his talk about self-experimentation.

Yet, to pigeonhole Seth as a primarily a self-experimentalist fails to understand what he was really about as a thinker.  As a professional research psychologist, he focused on developing “productive explanations” — explanations that help us not only to make sense of an individual surprising observation, but that also make predictions about additional diverse and often novel practical applications.  Self-experimentation was one important input, but certainly not his only source of experimental material. Self-experimentation has the virtue of allowing one to do more experiments in less time without spending a lot of money.  Self experiments allow you to make incremental progress rapidly, to adjust and learn quickly.

But Seth didn’t stop there  He typically synthesized results from many different fields into a coherent explanation.  He was not against using data from larger experiments, even controlled double-blind experiments.  It’s just that large “designed” experiments sometimes become unwieldy and expensive failures.  Simple self-experiments get you started “learning by doing” and often allow you to make rapid progress and weed out untenable hypotheses quickly, before you sink a lot of time and effort into your investigations. The best example of how Seth combined self-experiment with classical science may be how he came up with the Shangri-La Diet, a seemingly wacky–but actually very effective–way of losing weight, safely and without hunger.

In remembrance of Seth, I’d like to share my thoughts on that diet and Theory behind it–and the big impact it had on the development of my perspective on health and metabolism that is represented by this blog, Getting Stronger. As elaborated in Seth’s book, The Shangri-La Diet involves consuming small doses of “flavorless calories” between meals.  Typically this involves swallowing one or two tablespoons of unflavored sugar water or oil (e.g. olive oil or coconut oil) at least an hour before or after otherwise normal meals.  Otherwise, there’s no need to change the make-up of your normal diet. The result is a dramatic decrease in appetite, resulting in effortless weight loss.  As Seth tells the story, the idea for this weight loss method first came to him from an odd personal experience. While staying in Paris, he happened to consume some European sodas with flavors he had never before encountered.  Strangely, his appetite disappeared–even though he likes French food–and he lost a lot of weight.  He later found that plain, unflavored fructose water and unflavored oil–such as ELOO (Extra Light Olive Oil)–had the same weight loss effect. He started to experiment with the idea of eating foods with unusual and unfamiliar flavors and similarly lost weight. Seth found that merely consuming foods with bland or very little flavor also caused weight loss.

Others who tried the diet were amazed at how effective it was.  An entire internet forum captures countless experiences and tips provided by those who follow the diet. Pure “self-experimentation” might have just left things with the simple observation that you can lose weight by eating foods with unusual flavors or bland flavors. But Seth was as much a theorist as he was an experimentalist. He wanted to know why he lost weight by imbibing unflavored sugar or oil.  As a professor of psychology at UC Berkeley, he was trained in the field of classical Pavlovian conditioning, working with rats. So he revisited several independent streams of research and uncovered obscure experimental papers that might help him figure what was going on with the French sodas. As he wrote in the introduction to his book:

I think my secret weapon was that I combined three methods of investigation that had not been combined before. One was my scientific training, which made it easy for me to understand the relevant research literature.  Another method I used was self-experimentation: I wanted to lose weight and tried many different ways of doing so.  The third method was to play reporter: to phone weight-control experts and ask them about their research….By themselves, none of these methods is unusual, but they are rarely, if ever, combined. (SLD, p. 2)

A new theory of weight loss. The original step that Seth took was to generalize from his personal observations by combining them with research from two unrelated fields to devise a new theory of weight loss based upon the idea of flavor-calorie association:

The theory behind the Shangri-La Diet is based upon research in two areas that are usually separate: weight control and associative learning. Most weight control researchers know little about associative learning, and most associative learning researchers know little about weight control.  The two topics are studied in different university departments.(SLD, p. 142)

Specifically, Seth combined findings from four different researchers, much of it based upon investigations with laboratory rats:

• Gordon Kennedy’s research supporting the idea of body fat “set points”
• Michael Cabanac’s research on homeostatic regulation of appetite and “pleasantness” of foods
• Anthony Sclafini’s research on flavor-calorie association and food preference
• Israel Ramirez’s research showing how flavor alone could dramatically change the fattening properties rat chow, without any change in caloric content.

The result was a new theory of weight loss that synthesized Seth’s own self-experimentation with the published research.  The theory can be summarized in a single sentence:

Foods with a weak flavor-calorie association will lower your body fat set point and cause you to lose weight; foods with a strong flavor-calorie association will do the opposite.

Seth’s novel theory provides an explanation for why he lost weight drinking French sodas:  Despite the sugar calories in the soda, the flavors were unfamiliar, so Seth had not yet formed a psychological association between those flavors and the calories in the sodas. As the theory would predict, his set point dropped.  Because his weight now exceeded his set point, he would continue to feel satiated until his weight dropped below the new set point.  But beyond this explanation, what really excited Seth was that the theory provided predictions for new ways to lose weight:

No one believes a theory but the theorist, the saying goes. It’s not entirely true. Two sorts of findings are especially persuasive. First, confirmation of a surprising prediction. A famous example of this is the return of Halley’s Comet at the predicted time–very convincing evidence for the theory that comets orbit the sun.  One surprising prediction of this weight-control theory that has been confirmed is that extra-light olive oil will cause weight loss. The usual idea has been that consuming fat causes weight gain.  Second, repeated usefulness. When a theory is useful again and again, skepticism melts away. (SLD, p. 48-49)

A wellspring of new applications.  Chapter 6 of his book details how Seth’s flavor-calorie association theory suggested six different ways to lose weight.

• try new foods
• cook varied meals at home to avoid “ditto” fast foods (foods that are always the same, like Big Macs or Coke)
• add random flavors
• eat one food at a time in isolation (single flavors satiate more quickly that combined flavors)
• eat foods that are digested more slowly
• eat foods with less flavor

This is pretty cool because what Seth found was not a rigid dietary strategy, but a versatile framework for losing weight.  This framework provides more variety and freedom to experiment — something that a lot of diets overlook to their detriment, causing people to give up out of boredom.  The six methods that Seth listed in his book are not the end of the story.  Others picked up where he left off and discovered additional methods. One other method was to consume highly flavored non-caloric beverages or foods, such as herb teas or sugarless chewing gum, well before or after caloric meals or snacks. Another particularly effective  method, proposed by Timothy Beneke, is called “nose-clipping”: pinching your nose shut or wearing a nose clip while eating food or even the supposedly flavorless oil.  Beneke found that some people are particularly sensitive to flavors and can associated even relatively weak flavors with calories. Pinching your nose shut cuts out vitually all of the flavor.  If you don’t believe it try it–eat something very tasty with your nose pinched!   Flavor is largely aroma detected in the nose and not on the tongue. The tongue does detect a few simple sensations such as sweetness and saltiness, but apparently the brain does not associate these with calories.  Interestingly, at least anecdotally, overweight people are often those with the most pronounced sensitivity to flavor — so nose-clipping is particularly helpful for them.

I picked up my copy of the Shangri-La Diet book when it was published in 2006 as a best-seller, tried it, and easily lost 10 pounds.  I found it much easier that my other primary weight control method at the time:  low carb dieting.  The appetite suppression was incredible and liberating.  I would never consider taking diet pills, but I imagined this is what that would feel like.   It was so weird and counterintuitive that consuming caloric sugar or oil could have this weight loss effect.

A stimulus to new thinking.  But for me personally, the lasting impact of my encounter with the Shangri-La Diet and Seth Roberts was intellectual.  It was experimentation with the Shangri-La Diet, and later reading of Gary Taubes’ Good Calories, Bad Calories that really got me thinking about the physiological mechanisms of weight loss and weight regulation.  Both books provide evidence that the conventional account “calories in, calories out” cannot be the whole story.  Of course, nobody doubts the laws of thermodynamics.  But the body is not a simple calorimeter; it is a highly regulated homeostatic system, responsive to the signaling of hormones and neurotransmitters.  The sugars, fats, proteins and micronutrients in our food do more than provide fuel for energy; they also break down into substances that interact with receptors in the gut, liver, and muscle tissue.  They modulate hormone levels,  and interact with the hypothalamus — the brain’s control center for appetite, thirst, sex drive and thermoregulation. This intrigued me, and so began a lengthy correspondence with Seth, both via email and through posts on his “SethRoberts.net” discussion forum boards.  Inspired by Seth’s model of combining self-experimentation with experimental science and theory from diverse fields, I began a years-long exploratory search for the mechanisms that underly weight control.

While I believe Seth’s flavor-calorie association /set point theory represents a major step forward, it left me unsatisfied at the level of simple physiology.  How does the body fat “set point” work and are we all born with a genetically determined set point?  If so, why are some people able to lose huge amounts of weight and keep it off, where others fail?  What is really going on? So on September 21, 2006, under the moniker “NTB” (my initials),  I started a thread on the SethRoberts.net chat boards raising the question of whether the success of the Shangri-La Diet could better be explained by concepts other than “set point”.  The thread, entitled “Appetite Suppression vs. Setpoint Adjustment“, took many twists and turns over several years and I believe remains the most active single thread on his site, with more than 400 written responses and over 76,000 views.  After an initial exchange with Seth, I disappeared, but came back on April 19, 2009 with several posts that ignited a long series of interesting exchanges, ultimately culminating in a very different explanation of why the Shangri-La Diet works.  In addition to the extended theorizing in that large discussion thread, I posted other self-experiments on his discussion boards, including an interesting set of experiments on using a Blood glucose monitor as a weight loss tool.

Rewiring our reactions. Based upon the glucostatic weight control theory of Jean Mayer, I think we can discard the idea of a built in “lipostat” or body fat “set point” in flavor of an alternative concept based on blood glucose homeostasis and response of insulin and other hormones to not just flavors but also other psychological “triggers”.   I also think the lipostatic “set point” construct is at odds with a lot of evidence suggesting that our metabolisms are highly adaptive. Changing how we respond to triggers such as flavor can change our metabolisms.  Experimental evidence and my self-experimentation indicate that body fat levels can be changed significantly, without the need to carefully control flavor intake — if we “decondition” the way we respond to flavor using essentially the same techniques the Pavlov used a century ago to get dogs to salivate in response to a bell.  We can enjoy flavor without it being problematic for weight gain. My proposed alternative explanation for the effectiveness of SLD and three other related diets is based on the glucostatic theory and the conditioned responsiveness of signaling by insulin, ghrelin and other appetite hormones to flavor and other triggers. This alternative  is laid out in some detail in a 2011 blog post on Flavor control diets.    It’s not just alternative explanation for the sake of being different.  As Seth emphasized, the adequacy of a theory is not just how well it explains a few immediate facts, but whether it makes new predictions that turn out to be valid.  As detailed in the above blog post, the insulin-regulated glucostatic theory actually makes a number of different predictions that  are either not predicted by the flavor-calorie association theory or are diametrically opposed to what it would predict.  For example,

1. Consuming pure fats like olive oil or heavy cream should suppress appetite even if they are flavored. This only works, however, if no more than a trace of carbohydrates or proteins is present. It works because fats are noninsulinogenic.
2. Increasing the dose size of oils even to large doses should increase appetite suppression
3. Increasing the dose size of sucrose or glucose beyond the minimum dose, should reduce appetite suppression.  A small amount of sucrose or glucose increases satiety because it raises blood sugar slightly, and flies “under the insulin radar”.  Insulin is not secreted until blood sugar rises above a certain threshold, typically 120 mg/dL or so. But once it exceeds that threshold, insulin kicks in, blood sugar drops, and appetite increases. (Eventually, insulin does suppress appetite once it reaches the hypothalamus).
4. Sipping sucrose or glucose slowly will maintain appetite suppression.  If this is done slowly enough, the addition to blood sugar just balances out the amount of blood sugar consumed to meet metabolic needs.  But this is a careful balance.
5. Increasing the dose of fructose, xylitol, erythritol, or other non-insulinogenic sugars should suppress appetite at any dose.

While several of these predictions (2, 4 and 5) are consistent with the flavor-calorie association theory, others (1 and 3) are contrary to it’s fundamental assumptions, and would not be predicted by it. Yet I personally found that consuming coffee with heavy cream, but no sugar, had the same appetite suppressing effect as flavorless olive  oil or cream without the coffee flavor.  Several of us on Seth’s forums tested and documented the success of this approach.  We used the term “platinum calories” to describe such flavorful drinks or foods that contain fats like cream, but have no or little carbohydrate or protein.  According to a strict interpretation of the Shangri-La Diet,  such an effect should not exist. Sustainable weight control. My issues with the Shangri-La Diet go beyond the question of why it works.  While I believe that it is an incredibly clever, simple and safe method of short term weight loss, and I periodically use it as a tool to help with intermittent fasting, I don’t think it provides a satisfying long-term way of eating.  It doesn’t effectively address the root problem of poor weight control, and thus it requires the continued use of low flavor foods or non-caloric flavors, ad infinitum.  I’d rather permanently change the way I respond to flavors and adopt a sustainable way to eat.  The adaptive character of metabolism continues to form one of the major themes in my blog, discussed in multiple posts such as these:

• The Deconditioning Diet
• Change your receptors, change your set point
• Does tasty food make us fat?
• Does insulin make you fat?
• Obesity starts in the brain
• Learning to fast

My debt of gratitude to Seth.  While my thinking on diet has diverged in key respects from the ideas set forth in The Shangri-La Diet, that book represented an intellectual watershed for me.  Seth set an inspiring example for how to harness personal experience and simple self-experiments as an engine of insight in order to generate new ideas. He taught us how to test these ideas by exploring their predicted applications, while integrating the results within the broader fabric of established scientific theory and experiment.   In doing so, his personalized approach empowers all of us to apply the scientific method as a fast, effective and inexpensive way to learn about ourselves.  Self experimentation is a method that doesn’t replace, but rather complements and sometimes challenges or extends, the findings of established science. Thank you, Seth, for helping us finding new ways to improve our health and well-being.