Metabolic Effects of a Traditional Asian High-carbohydrate Diet

A recent study supports the notion that an 'ancestral diet' focused around high-starch agricultural foods can cultivate leanness and metabolic health.

John McDougall gave Christopher Gardner a hard time at the McDougall Advanced Study Weekend.  Dr. Gardner conducts high-profile randomized controlled trials (RCTs) at Stanford to compare the effectiveness of a variety of diets for weight loss, cardiovascular and metabolic health.  The "A to Z Study", in which Atkins, Zone, Ornish, and LEARN diets were pitted against one another for one year, is one of his best-known trials (1).

Dr. McDougall asked a simple question: why haven't these trials evaluated the diet that has sustained the large majority of the world's population for the last several thousand years?  This is an agriculturalist or horticulturalist diet based around starchy foods such as grains, tubers, legumes, and plantains, and containing little fat or animal foods.  Researchers have studied a number of cultures eating this way, and have usually found them to be lean, with good cardiovascular and metabolic health.  Why not devote resources to studying this time-tested ancestral diet?  I think it's a fair question.

The reason seems to boil down to practicality: it's hard to get people to eat that way for a long period of time.  It's not easy to get people to adhere to a relatively bland 'third world diet', and the NIH isn't overly excited about studying a diet style that can't reasonably be applied to the general population of an affluent nation.

The Study

Still, we get scientific hints every now and then that this style of eating, if executed intelligently, is part of the spectrum of ancestral diets that can support good health in most people.  Such a study was recently published in PLoS One, which is an open-access journal so you can all have a look if you want (2).

William C. Hsu and colleagues randomly assigned 28 Asian-Americans and 22 Caucasian-Americans, all of which were deemed at risk of developing type 2 diabetes, to two groups:

  1. 16 weeks of a "typical Western diet" (TWD), presumably representative of the current US diet.  The methods provide very little information about what the TWD actually was, besides providing 50% of calories from carbohydrate, 16% from protein, 34% from fat, and 6 g of fiber per 1,000 kcal of food.
  2. 8 weeks of a TWD followed by 8 weeks of a "traditional Asian diet" (TAD).  The methods provide very little information about what the TAD actually was, besides providing 70% of calories from carbohydrate, 15% from protein, 15% from fat, and 15 g of fiber per 1,000 kcal of food.
Basically, it was a high-carb, low-fat, high-fiber diet vs. a more typical affluent diet.  The participants were fed enough of each diet to maintain body weight, based on estimated calorie requirements, and energy intake was deliberately adjusted to avoid weight loss.  The researchers provided all food by delivery during the 16-week study, but since participants weren't housed at the research facility, there's no way to be certain of exactly how well they complied.  However, compliance tends to be good when researchers provide all food like this (people like free food).

There were a number of dropouts in the first 8 weeks of the study, leaving 33 people in the intervention group and only 7 in the control group.  This unfortunate turn of events does weaken the study.

The Results

Despite the conscious attempt to prevent weight changes, participants gained weight and fat during the TWD phase and lost weight and fat during the TAD phase.

While eating the TAD, participants experienced a few metabolic changes that appear beneficial.  The amount of circulating insulin and glucose following a glucose challenge decreased, estimated insulin sensitivity increased, and LDL decreased (although so did HDL).  This is despite the fact that they used "intent-to-treat" analysis that lumps dropouts together with completers.

The participants of Asian descent reacted similarly to the diet as those of Caucasian descent.


This study had some weaknesses, due both to its unusual design and the unfortunate (and asymmetrical) number of dropouts.  However, it adds to the evidence that horticultural and agricultural diets centered around unrefined carbohydrate can benefit weight and health, relative to the typical affluent diet.

Some people have speculated that very low-fat, or very low-carbohydrate, diets may be able to cause weight loss independently of their calorie content.  This study could certainly add fuel to that fire.  However, due to the less-than-perfect control the researchers had over the subjects' diets, I'd be cautious about making that conclusion.  The best available evidence continues to suggest that the calorie value of food impacts body fatness, but macronutrient composition doesn't.  I believe we'll eventually learn that the story is a bit more complex than that, but the boring adage "a calorie is a calorie" is the interpretation the evidence currently supports.

Although obesity and cardiovascular/metabolic disease are rare in horticultural and agricultural societies, that doesn't mean these cultures are always healthy.  Some of them suffer from diet-related health conditions, often caused by nutrient deficiencies.  According to my understanding, these tend to result from one of three causes:
  • Low diet diversity.  Eating an excessively grain-heavy diet can lead to mineral and vitamin deficiency disorders, including mineral deficiency rickets, vitamin A deficiency, scurvy, anemia, and pellagra.  Eating starchy tubers to the exclusion of other foods can lead to protein and folate deficiency.  Solution: eat a diverse diet that includes animal foods, starchy tubers, vegetables, and nuts.  I speculate that vegetables and dairy are particularly valuable to agriculturalists because they complement the shortcomings of a grain-based diet.
  • Nutrient-poor soil.  Some soils are deficient in specific minerals, and this can translate to deficient food.  For example, iodine deficiency cretinism used to be common in the Alps due to iodine-poor soil, leading to endemic mental retardation.  Even though cretinism is rare in the Alps today, the term crétin des Alpes, or simply crétin, remains a common insult in France.  Parts of the Midwestern US also have iodine-poor soil, and iodine deficiency goiter was common there until the introduction of iodized salt.  Solution: eat diverse foods from a variety of places, which is the default today anyway.
  • Inappropriate food preparation.  Traditional preparation methods such as fermentation and nixtamalization are used throughout the world to increase the nutritional value of grains (3).  Cultures that don't apply these methods are more susceptible to deficiency diseases.  For example, in Iran, traditional agriculturalists who rely heavily on unfermented whole wheat develop mineral deficiency rickets and osteomalacia, despite spending much of the day in the sun; the same effect has also been noted in other populations  (4, 5).  The adoption of corn throughout the world (because of its incredible calorie productivity) was accompanied by a wave of pellagra, caused by niacin deficiency.  Native Americans who had been eating corn for thousands of years didn't suffer from pellagra because they pre-treated their corn with the mineral lime (calcium hydroxide) to unlock its niacin and make it a good source of calcium-- a process called nixtamalization-- but this knowledge didn't accompany the plant to other parts of the world.  Solution: use traditional grain preparation methods and/or eat a diet that doesn't rely too heavily on grains.
My opinion is that grains in general can be part of a diverse healthy diet, but very grain-heavy diets present nutritional challenges that must be managed.  I also believe that wheat is problematic for some people, for multiple reasons*, although many people probably tolerate it just fine.  There is no nutritional need for grains in the diet, including wheat, and their primary virtues are their low cost and good taste.

* 1) The gluten-provoked autoimmune disorder celiac disease affects nearly 1% of the US population, and some evidence suggests that the prevalence increases with age.  People with celiac disease have a higher mortality risk than the general population, and most cases remain undiagnosed.  That alone is a huge chronic disease burden attributable to a single food.  2) Although research is ongoing and the area remains controversial, some research suggests that certain people without celiac disease are nevertheless sensitive to gluten.  The true prevalence of non-celiac gluten sensitivity remains unknown.  3) Wheat is a major source of FODMAPs, a class of fermentable fibers that exacerbates the extremely common digestive condition irritable bowel syndrome (present in 10-20% of Americans).  Wheat is one of many dietary sources of FODMAPs, however it's one of the largest in US and European diets.  4) Wheat is usually consumed after having been ground into flour and baked, often in combination with fats, sugars, and other desirable ingredients.  This creates highly palatable, calorie-dense foods that lead to overconsumption.