Seeking Shangri-La

I’ve been reading The Shangri-La Diet by Seth Roberts, a psychology professor at UC Berkeley. I’ve mostly managed to stay away from fad diets over the years, but this one got some unusual endorsements from people I respect including Kathy Sierra and the Freakonomics duo.

The basic idea is that twice a day you eat one to two hundred calories worth of sugar water or extra light olive oil. You must do this at least one hour after and one hour before eating or drinking anything else except water. Roberts claims this reduces your “set point” and thus reduces your appetite. Consequently you eat less, and lose weight.

I’m skeptical. For a professional scientist, Roberts seems way too fond of anecdotal evidence. The Shangri-La Diet reads like a typical fad diet book, and seems lacking in serious evidence. At best it’s a hypothesis with some anecdotal evidence to support it. It’s suggestive, but nothing more.

I’d like to see some serious controlled double-blind studies. As nutrition studies go, this one should actually be fairly easy. First of all, the diet doesn’t require much (or any) willpower on the part of the participants. One failure of a lot of existing nutrition research is that the participants cheat. They don’t eat what they’re supposed to, and they do eat what they aren’t supposed to. Then they lie about it to the researchers. Normally the only way to do reliable research on eating is to control people’s food intake 24 hours a day, seven days a week; and that’s rarely feasible.

Secondly, it’s really hard to do double blind studies because everyone can tell what food they’re eating. For example, if you want to find out if people who eat a grapefruit a day lose more weight than people who don’t, every participant knows whether or not they’re eating the grapefruit. In this case, though, you could easily give group A sugar water and group B non-calorific water sweetened with saccharine or aspartame or some such.

In other words, studies that meet the highest standards of scientific research are possible here. I’d like to see some.

Nonetheless, the diet is:

  1. Relatively easy to perform
  2. Maintainable over a long period of time
  3. Cheap
  4. Doesn’t seem likely to do any serious harm, even if it doesn’t work

So I may give it a try and see what happens.

2 Responses to “Seeking Shangri-La”

  1. ken Says:


    The research has been done thousands of times and the results are always same.
    In engineering terms, treat your body like a control volume: if the amount of calories you consume is smaller than the number you burn through exercise, you will lose weight.

    Eat salads, whole grains, lots of vegetables, a modicum of poultry/fish.
    Exercise portion control -eat til you’re a bit shy of satsified but not profoundly hungry.
    Make a exercise a top priority & do it 30 minutes a minimum of four times a week.

    All diets depreciate the quality of your health. The body is a complex organism that needs vitamins and minerals and it needs to get them from food, not pills.

    Thanks & keep up the great tech writing

  2. Elliotte Rusty Harold Says:

    The research has not been done on the specific claims Roberts makes. (At least not on humans; some research has been done on rats.) Nothing he says denies the basic point that you need to eat fewer calories than you expend to lose weight. As you say, we’ve known this for a long time.

    The question then becomes how does one motivate people to eat fewer calories? Simply telling people to eat less and exercise more is clearly not working. Some things do work. Fen-phen was a marvelous appetite suppressant that helped many people lose huge amounts of weight. Unfortunately, it was toxic and could not be sustained over a period of years.

    Roberts’s claim, which I am not vouching for, is that intake of a few hundred calories of flavorless foods sufficiently far away from flavored foods, will make people feel fuller, think about food less, and eat less. In essence it is a realtively natural equivalent of fen-phen. His hypothesis is fully consistent with what is generally known about weight loss. It does not contradict the standard model of how the body gains and loses fat in any way.

    His essential claim is that he can trick the body into eating less. Most diets clearly fail at this over the long term. Most diets are also obviously unsustainable over a lifetime. Low-carb diets are simply the latest example of this. There are numerous other examples.

    The Shangri-La diet does seem at least plausible to sustain for a lifetime. It’s not really a diet at all. All it involves is drinking some sugar water or extra light olive oil a couple of times a day. Then eat what you feel like eating. Roberts claims that what you feel like eating under this regimen is a lot less than what you would otherwise feel like eating.

    Again, whether this is true or not, I don’t know; but it’s not as obviously false as some of the fads one encounters; and it’s easy enough to test. There’s a definitely a strong whiff of pseudo-science and quacksterism to the book; but not nearly at the same level as you find in the works of the diploma mill-Ph.D’s and fake doctors who infest the field. There just might be something real here; but I’d really like to see some serious double-blind studies to settle the question definitively.

Leave a Reply