Chris Masterjohn has put up a detailed review of our book. He calls my wife and I “awesome,” the book “excellent,” and awards “two thumbs up.” Thanks Chris!

Chris’s review led me to reflect a bit on our book, and I thought it might be appropriate to close the blogging year with a couple of posts about it. Specifically, I will:

1. Clarify our argument for optimal macronutrient ratios. Chris spends a substantial part of his review discussing the issue of carbs, and I think readers may be interested in my perspective on the issues he raises.
2. Look at the “big picture”. What is the state of dietary science, and where does it need to go? How close are bloggers and scientists to integrating the various perspectives and reaching the truth? What does our book in particular contribute to both popular and scientific debates?

Today I’ll discuss the carb issue.

Macronutrient Ratios and Carbohydrate Intake

Step One of our book is about macronutrient ratios. The key goal of Step One is to help readers understand how much of each macronutrient they need to avoid deficiencies and what levels will produce toxicity conditions.

For some macronutrients this is easy, because the evidence is fairly clear:

• Protein. Roughly 5% of energy as protein is sufficient to avert a deficiency, and the evidence of toxicity is also fairly clear: protein over 40% of energy can be fatal, and there are less severe indications of toxicity at about 30% for adults, 20% for pregnant women, and 9% for infants.
• Polyunsaturated fats. About 0.5% of energy will avert deficiencies of both omega-6 and omega-3 fats, and there are subtle indications of toxicity above about 4% of energy, with the indications becoming obvious above 10%-15% of energy.
• Saturated and monounsaturated fats. These are benign at all intake levels: non-toxic in high doses, and impossible to become deficient in (except during starvation) because the body will manufacture them from carbohydrate or protein.

But determining the optimal intake range of carbohydrates is difficult, because unlike the other macronutrients there is no data that lacks confounding factors.

The main problem is that the main sources of carbohydrates in western diets – grains and sugar – contain food toxins such as gluten and fructose. So any benefits from reducing carbs could be the result of reduced toxin intake, not reduced carbohydrate calories.

We know these plant toxins are important, because populations that obtain carbs from “safe starches,” such as yam-eating Kitavans, rice-eating Chinese and Japanese, and potato-eating Irish have been notably healthy. (See, for instance, our China Study posts, our wheat series, and Stephan’s Kitava and potato posts.)

However, in none of those “safe starch” eating populations do we have good data on how health outcomes vary with a differing mix of carbohydrates vs. saturated and monounsaturated fat.

With data lacking, our prescriptions have to be based largely on biologically-informed theory – theory instructed by, as Chris puts it, “circumstantial evidence.”

The “Eat What We Are” Argument

A good first estimate for carb intake comes from the simple precept that we should “eat what we need.” The body’s glucose utilization level, about 600 calories per day, tells us how much we need. The simple prescription, then, would be to eat about 600 calories, or 30% of energy for a slender person, as carbs. Let’s call this intake a “moderate carb” diet.

Evolutionary Arguments

Another place to look for guidance is evolution: what macronutrient ratios has evolution selected for?  Here we do not place much weight on ancestral Paleolithic diets, which varied substantially from place to place and time to time. Rather, we place more weight on animal diets and the composition of mother’s milk.

Both human and animal breast milk is majority-fat, moderate-carb, low-protein. In milk of all species, the carb percentage closely tracks the infant’s glucose needs.

In animal diets, the macronutrient ratios delivered to the body depend on the design of the digestive tract, which transforms the macronutrient content of food. Digestive tracts are generally good at transforming carbs into fat. As a result, after transformation most animals have carb intakes equal to or below their carb needs. In ruminants and foregut fermenters, the transformation of carbs into fats is virtually complete; in hindgut fermenters, it is only partial.

Many animals achieve so many of their calories from bacterial fermentation of dietary carbs that their diets are actually highly “ketogenic” and nearly zero-carb.

Thus, evolutionary considerations suggest:

• Optimal human carb intake is probably close to or below our glucose utilization of 600 calories per day.
• Ketogenic diets with substantial calories from short-chain fats are probably healthy.

I find evolutionary logic convincing. Since healthy creatures will generally have more descendants than unhealthy ones, evolution surely selects for health.

The Argument From Palatability

In addition to being healthy, a diet should be tasty – palatable. People should enjoy it and be pleased to eat it.

This has an influence on macronutrient ratios. Fatty animal foods in general have higher caloric density than carb-rich plant foods. Fruits and berries have 200 calories per pound; potatoes and sweet potatoes 300; taro 600; white rice 1300. In contrast, meat often has about 1000 calories per pound; bacon has about 2400. To eat 30% of calories from carbs, while avoiding toxic foods like grains and sugar, requires eating substantially more plant food than animal food.

But people generally find food most palatable when there is a mix of plant and animal foods. Since carbs come exclusively from plant foods (excepting milk), which have lower caloric density, this means a tasty diet inevitably has more fat than carb calories.

The Argument From Toxicity

As we note in the book, plant foods have more toxins than animal foods. This is because they use toxins to poison animals that might eat them, while animals cannot easily poison predators without also poisoning themselves.

Toxins are found in nearly all plant foods. Peter at Hyperlipid has a post today about berry toxins, pointing out that they have a similar biological effect to a pharamaceutical drug, anacetrapib, that raised the death rate in clinical trials. (I note in the comments that the benefits from poisoning gut pathogens and cancer cells might outweigh the damage from direct toxicity. This is why I eat “Neo-Agutak” – cranberries. )

If it’s desirable to avoid toxins, it’s desirable to keep plant food intake down and animal food intake up. Given the low caloric density of most plant foods, that also implies that carb calories will need to be low on an optimal diet.

Plateau Ranges

As Chris notes, we do not define precise optima for each nutrient, but rather suggest “plateau ranges” that achieve all the benefits and none of the toxic effects.

Based on the above reasoning, for carbohydrates, our estimated plateau range is from 10% of energy on the low end (which we recommend for therapeutic ketogenic diets) to 30% on the high end (a “moderate carb” diet in which all the body’s glucose needs are met from diet).

We do suggest specific values of 20% carbs, 15% protein, and 65% fat. These are mid-range values which casual dieters may wish to aim for; even if they miss by 10% of energy on any macronutrient, they will still be in the plateau range.

What Happens If You Deviate From the Optimum?

But if this logic tells us what the optimal carb intake is, it does not tell us how much harm is done by carb intakes different from the optimum.   

Here I agree with Chris that in healthy people, the damage to health from changes in the carb:fat ratio could be very small, as long as the carb sources do not contain appreciable toxins. The body has a great capacity to convert carbs to fat. So long as that transformational capability remains undamaged, a very high-carb diet may be functionally very little different from an optimal diet.

In metabolically damaged people, however, transformations from one macronutrient to another may be impaired. In this case, the body may have very little flexibility in repairing a suboptimal carb:fat ratio.

I thus agree with Chris’s conclusion:

I thus find it very difficult to believe that humans must maintain their diets within a very narrow range of macronutrient ratios. More likely, the human body is extraordinarily flexible and can tolerate a wide range of carbohydrate and fat intakes.

That said, it may well be the case that in modern folks with damaged metabolism, low-carb is the way to go. Indeed, well controlled studies have clearly shown that low-carbohydrate diets are superior to low-fat diets by a multitude of criteria for people with various markers of the metabolic syndrome (references here, here, and here).

Nothing in the book contradicts this conclusion. As an editorial decision, we chose to describe the “perfect” diet for health without spending space discussing imperfect diets. Given that decision, we didn’t need to address whether a 90% carb diet is only marginally worse, or much worse, than a 30% carb diet.

But that is an appropriate topic for discussion on the blog. I’ve begun a “zero-carb dangers” series; perhaps I should do a “high-carb dangers” series too.

Conclusion

Much research will be needed to disentangle the relative contributions of plant toxins and carbohydrate calories to health.

It looks from early evidence that intake of plant toxins may be much more consequential for health than the overall carbohydrate level. This is good news, since selection of low-toxicity foods and preparation methods that detoxify foods may enable big improvements in human health.

Nevertheless, even if most toxins can be removed, there are reasons to believe that a low- to moderate-carb (10% to 30% of calories as carbs) diet will be optimal for human health. Such a diet seems to work well for nearly everyone, even the metabolically damaged; and it produces a tasty diet that is a pleasure to eat. It therefore seems the most prudent recommendation for diet book authors to make. And we stand by it!