Search This Blog


Wednesday, March 21, 2012

Yet Another Silly Study About White Rice!

If you look at articles about East Asian countries in issues of National Geographic from the early 20th century, you will notice two things. One is that many of the people in East Asia were eating a lot of white rice. The other is that only the rich people and the sumo wrestlers were overweight. That’s because the rich people and the sumo wrestlers were eating something besides rice and vegetables. Back in the 1930s, Dr. Walter Kempner realized that populations that ate a rice-based diet were slim and remarkably free of degenerative disease. After he started teaching his patients at Duke University in North Carolina in the United States to eat a diet consisting of white rice and fruit, they lost weight and reversed their type 2 diabetes. Many of his patients with type 1 diabetes were even able to cut their insulin doses. So why am I seeing studies that are trying to prove that white rice causes diabetes? The latest was just published in the British Medical Journal (BMJ).

One problem with the BMJ study is that it analyzed how much white rice people were eating, but it didn’t account for how much food they were eating and what other kinds of food they were eating. It just looked at a dose effect between servings of rice and health outcomes. That approach is potentially misleading. For example, someone who is in training to become a sumo wrestler probably eats a lot more rice than the average Japanese person does. However, he also eats a lot more meat and drinks a lot more beer. Because of their training diet, sumo wrestlers gain weight and often get type 2 diabetes. Yet it would be extremely difficult for someone to gain enough weight to become a sumo wrestler or even stay overweight on a diet of white rice and vegetables. It would therefore be silly to blame white rice for a sumo wrestler’s weight and blood sugar problems.

The BMJ study provides a classic example of a recurring problem I see in the scientific literature on nutrition. The researchers want to answer a question—in this case, whether white rice makes people fat and contributes to type 2 diabetes. The sensible way to answer that question would be to survey all of the available evidence and then see how it relates to Hill’s considerations for establishing cause and effect in epidemiology. Instead, the BMJ study took a very biased view of only one kind of evidence and then did some fancy math. Then, they came up with a misleading result.

The researchers wanted to know whether consumption of white rice is linked to obesity and diabetes. So they gathered a bunch of studies that asked people about their rice consumption and that followed up to see which people gained weight and got type 2 diabetes. Then the researchers used a mathematical approach called meta-analysis to combine the results of the studies. The biggest problem with this approach is that the researchers chose only one kind of study for their analysis: cohort studies of people who weren’t diabetic at baseline and who were eating whatever they felt like eating. This means that the researchers systematically left out studies that compared rice-eating populations to other populations, as well as the clinical studies of people who lost weight and reversed their diabetes after being taught to eat as much as they wanted of a diet based heavily on rice and vegetables.

We’ve known since the early 20th century that a high-fat diet promotes insulin resistance, which is the underlying problem in type 2 diabetes. In the early 1930s, a British researcher named H.P. Himsworth found that he could cause insulin resistance in healthy young men in a week by feeding them a diet that was 80% fat by calorie. As he replaced fat with starch in his test diet, his subjects’ glucose tolerance improved. He got the best glucose tolerance with the starchiest diet he tested: 80% carbohydrate by calorie.

Starting in 1939, Dr. Walter Kempner found that he got great results by teaching his patients to eat a diet based on white rice, fruit, and fruit juice--plus some added sugar for patients who were losing too much weight on that low-fat diet. Kempner designed this extremely low-fat, low-salt, low-protein diet because many of his patients had kidney problems and high blood pressure. That was back in the days before blood pressure medications. The high carbohydrate content of this diet (>90% by calorie) and correspondingly low fat content helped to improve his patients’ sensitivity to insulin. The Rice Diet Program that Dr. Kempner founded is still helping people lose weight and reverse their chronic illnesses, but now they do emphasize whole grains.


  1. So you're naturally skinny, and you're metabolically inclined toward vegetarianism. Lucky you! Obviously not everyone is so lucky. Equally obviously, we all want to be thin, and we want it really really badly. It does no good, and considerable harm, for you to shout from the rooftops, "Look at me and do as I do," because it doesn't work. If it did, we'd all be as thin and smug as you are. You won the genetic lottery. Be happy and shut up.

    FWIW, sumo wrestlers eat a low-fat, high vegetable, moderate protein diet. The real source of their calories? Beer and rice -- 12,000 calories of rice per day. High GI carbs into a particular kind of body are both addictive and obscenely fattening. If you're immune, well bully for you...

    Also FWIW, I live in Chapel Hill, NC, and I attended Duke University in the 1990s. I know 3 people who did the Duke rice diet at the clinic. It worked for none of them. They invested considerable time, considerable effort, and considerable money, and they all ended up poorer, fatter and more despondent than before they started.

    Genetically fortunate, ideologically smug people like you sicken me, because you have _no idea_ what the rest of us are going through.

  2. Where do bison get their protein? How about mice? Raccoons? Tigers? Your point about gorillas is ultimately irrelevant to humans, because different species have different GI systems. Feed grain to a lion and you get a sick, then dead, tiger. A bison won't eat meat at all.

    Chimps and bonobos, of course, love meat, and eat it whenever they can get it. Have you seen the canines on a baboon? Humans, according to the best available science, evolved specifically to be high-heat daytime hunters, an ecological niche otherwise unfilled on the African veldt. In temperate zones that role is filled by wolves, but until the emergence of human-like creatures, it hadn't been filled in the hot, dry climate of sub-saharan Africa. Our lack of fur, bipedalism, and extreme physical endurance all made us superb daytime predators, and that was our original ecological niche. That we turned out to be extraordinarily flexible ecologically and nutritionally appears to be something of a serendipitous accident. But to make conclusions about humans from the nutritional requirements of gorillas is ridiculous.

  3. The gorilla gastrointestinal tract is very similar to that of a human being. If a gorilla can get enough protein from a plant-based diet to grow to be 400 pounds or more, why should I imagine that human beings would have trouble getting enough protein from a plant-based diet. In fact, a review of the published literature on human nutrition shows that human beings don't need to worry about getting enough protein if they are getting enough calories from any reasonable plant-based diet. The scientists who designed the food rationing systems in WWI knew that.

    The canines on a baboon are not for predation. Gorillas also have fierce looking canine teeth.

    It's silly to consider daytime hunting on the veldt to be a separate ecological niche from nighttime hunting on the veldt. The same prey animals that are on the veldt in the daytime are also there at night.

    1. The same prey animals are there, yes, but no one else is actively hunting them during the heat of the day. Humans are uniquely well adapted to physical endurance and heat management. Lions, for example hunt in the corners of the day, and at night. In the heat of the day, they rest, as do most of the prey animals. Humans, despite a bevy of physical disadvantages (no speed, no fangs, no talons, very little physical strength), quickly became dominant predators because we could run prey animals to exhaustion in the heat of the day, and because we were tool users and (later) tool makers. We overcame our lack of strength and carnivorous dentition through the discovery/development of cooking. There's been some really fascinating research on this lately that you might actually find interesting.

      To me, the most fascinating aspect of human evolution is the connection between the development of our brain and the loss of physical strength and fangs. As you probably know, pound for pound a chimp or a gorilla is about twice as strong as a human. This is, in short, because the myosin fibers in a non-human ape's muscles are considerably longer than the myosin fibers in a human muscle. As far as anyone can tell so far, this appears to have been a random mutation, but its impact on our evolution, and our ultimate ecological niche, was staggering beyond belief.

      To put it in its most condense form, weak jaw muscles allowed our craniums to grow while shrinking our jaws and teeth (I urge you to look this up. I'm being too brief, and this affects overall accuracy. The conclusion is sound though -- weak jaw muscles made big brains possible at the expense of big gorilla fangs.)

    2. Now, onto comparative anatomy and the digestive tracts of gorillas and humans. A gorilla and a human have somewhat related gut specifications, but the similarity is not so great as you report. A gorilla has a much larger gut (as measured both by weight, volume, and energy consumption) per pound of body weight than does a human (I've looked this up in the past, but can't provide a reference at the moment -- we can argue the point if you like ;-) ). A quick comparison of rib cages shows how much more space a gorilla's GI tract requires as a percentage of body space). There is reasonably good evidence that the human gut shrank to accommodate true bipedalism. A a giant grazer's gut is simply too large allow for a vertically balanced body (and the prevalence of of pelvic hernias among the obese is evidence that even now we're pretty close to the limit of what is anatomically possible).

      I have no problem with the concept that our GI system evolved from that of a frugivore, and I'll accept that it still shares some of its gross properties (length and tortuosity) with that of frugivores. But it is smaller, and has less absorptive area per lb. of body weight) than true frugivores, and yet has to provide sufficient nutrition to support our extremely energy-hungry brains. How can it do this? More energy-rich foods like meat and fat.

      Again, I'll accept that our GI tracts are transitional (or hybrid if you prefer). They have aspects of a frugivorous past, but their overall lack of size makes it very difficult, in the absence of energy rich grains (which have really only been a significant source of calories for humans for the last 10,00 - 15,000 years -- the blink of an evolutionary eye) to process enough energy to thrive. You'll probably disagree with this, but very few people stay on a truly fruitarian diet for more than a decade, and to my knowledge no young children instinctively refuse meat. Gorillas, as you know, are actually quite fond of insects (especially ants and termites).

      (I hope you don't see these posts as trolling. That's not my intention at all. I do disagree with much of what you say, but not at all disrespectfully. My own hypothesis about human nutrition is fairly complex, and I believe there are probably good reasons why some people thrive on vegetarian diets and others do not. It's a very long story, though, and I've blathered on long enough (without editing, since the window I'm typing in is about 4 lines long). Cheers.

    3. Actually, I do see your posts as trolling. You don't know what you are talking about. The fact that human protein needs are modest and easily met by any practical plant-based diet has been well established scientifically for more than 100 years.

  4. Sumo wrestlers do not eat a low-fat diet. They eat a lot of a dish called chanko-nabe, which is a stew containing chicken, fried fish, beef, or tofu.

  5. Chanko-nabe is in fact reasonably low fat, and it accounts for well less than half the calories consumed by a sumo during a normal training day. The bulk of the calories come from rice and beer. It is a vegetable and meat stew that is both high in fiber and protein. It is not particularly high in fat. Again, well more than half the calories in a sumo's diet comes from rice and beer.

    1. Chanko-nabe isn't low in fat!