Caution: Contains Science
© J. Stanton
In previous installments, we've proven the following:
  • A calorie is not a calorie when you eat it at a different time of day.
  • A calorie is not a calorie when you eat it in a differently processed form.
  • A calorie is not a calorie when you eat it as a wholly different food.
  • A calorie is not a calorie when you eat it as protein, instead of carbohydrate or fat.
  • A calorie is not a calorie when you change the type of fat, or when you substitute it for sugar.
  • Controlled weight-loss studies do not produce results consistent with "calorie math".
  • Even if all calories were equal (and we've proven they're not), the errors in estimating our true "calorie" intake exceed the changes calculated by the 3500-calorie rule ("calorie math") by approximately two orders of magnitude.
(This is a multi-part series. Return to Part I, Part II, Part III, Part IV, Part V, or Part VI.)

Empirical Evidence: A Calorie Is Not A Calorie When You Add Carbohydrate To A Zero-Carb Diet

There are many anecdotal reports of people finding it difficult or impossible to gain weight on a zero-carb diet, even with massive overfeeding. Yet there are controlled trials that seem to show high-fat diets having no such overfeeding advantage. Why not?

This study provides some clues:
J Nutr Biochem. 2003 Jan;14(1):32-9.
Effects of dietary carbohydrate on the development of obesity in heterozygous Zucker rats.
Morris KL, Namey TC, Zemel MB.

"...We fed 6-week old male heterozygous (fa/+) lean rats carbohydrate-free diets containing primarily saturated fat either ad libitum or pair-fed. These diets were compared to standard chow and to a high saturated fat mixed diet containing 10% energy from sucrose for 4 weeks."
This is a good start: many "high-fat" diet trials use industrial lard containing ~20% linoleic acid (an omega-6 fat), or industrial seed oils with even greater LA content - which, as we've seen in the previous installment, is strongly implicated in the development of obesity. Furthermore, most "high-fat" laboratory diets contain about 20% purified sugar...which, as we've previously noted, seems to be obesogenic by itself.
A Short Digression: "High-Fat" Almost Always Means High Sugar

The results of Morris 2003 call into question all obesity research featuring "high-fat diets". First, they're usually on a strain of mice (C57BL/6, or "black six") specifically selected for its propensity to quickly become obese when fed high-fat diets, unlike other strains of mice (let alone other animals, like humans, that aren't natural seed-eating herbivores). More importantly, in nearly every case, they use D12492 - a mix of purified ingredients containing no actual food, and specifically designed to make C57BL/6 mice obese as quickly as possible. (More here, via Dr. Chris Masterjohn)

D12492 contains 20% purified sugars.

So be skeptical whenever you see a headline claiming negative effects for "high-fat diets".
All diets were standardized to 20% protein and 11% corn oil. (Yes, these are industrial Frankendiets.) The carbohydrate-free diet contained 69% coconut oil; the 10% sucrose diet contained 59% coconut oil and 10% sucrose (table sugar), with no other carbohydrate; and the "standard chow" diet contained 59% cornstarch and no coconut oil.

Does that matter? 10% carbohydrate is still VLC, right? That's only 50 carbs on a 2000-calorie diet...and it's still almost 60% coconut oil, so they should be mostly in ketosis, right?

Here's what happened after four weeks. First, the food intake figures, from Table 2:

table 2
© unknown
The zero-carb rats ate 36% more "calories" than the chow rats, and about the same (3% more) as the 10% sucrose rats...and the pair-fed zero-carb rats ate the same number of "calories" as the 70% carb ("standard chow") rats. (That's what "pair-fed" means: one group is fed only as much as another group eats.) So if the CICO zealots are correct (IT'S PHYSICS!!!1!!1), the standard chow and pair-fed rats ought to be lean, while the zero-carb and 10% carb rats ought to be obese.

Meanwhile, back in reality, here's what happened:

Table
© unknown
Grams of fat gained during the 4-week feeding period, by diet.
"Weight gain was negligible in the carbohydrate free groups compared to standard diet and 10% sucrose diet (p = 0.03). This was reflected in energy efficiency which was markedly reduced (90%) in the carbohydrate-free groups compared to the other groups (p = 0.04)."
[...]
"Animals consuming the standard or mixed (10 en% sucrose) diets gained 90% more weight (p = 0.03) than animals consuming the carbohydrate-free diet ad libitum (Fig. 1A)." -Ibid.
The data is presented confusingly, so I'll put it all together in table form.
There are a couple problems with the data presentation in this paper. First, the "feed efficiency" graph doesn't appear to agree with the primary data they present, so I've used the values from Table 2 and Figure 1A to recalculate it. Second, the paper doesn't give the actual weight of the rats, just the change in weight...but given the typical developmental schedule of a heterozygous Zucker rat, it's likely between 200g and 400g.


Dietary group Dietary energy consumed
(1 kJ = ~4.2 dietary calories, or kcal)
% increase
in dietary energy
from baseline
% carbohydrate
in diet
Weight gain
per rat (g)
Weight gain
per megajoule
of energy consumed (g/MJ)
Standard (10% sucrose, 70% total carb) 22400 kJ 0% 70% 20 g 0.89
Zero-carbohydrate pair-fed 22700 kJ 1% 0% -1 g -0.04
10% sucrose, 10% total carb 29500 kJ 32% 10% 20 g 0.68
Zero-carbohydrate 30500 kJ 36% 0% 2 g 0.07


Let's put these results in English:
  • The 10% sucrose rats ate 32% more food than the standard (70% carb) rats, but gained the same amount of weight (20g).
  • The zero-carb rats ate slightly more than the 10% sucrose rats, and 36% more than the standard rats - but gained an insignificant amount of weight (2g).
  • The pair-fed zero-carb rats ate the same amount as the standard rats (that's what "pair-fed" means), but lost a negligible amount of weight (-1g).
Clearly, in this study, weight gain (and loss) doesn't correspond at all to "calories" consumed! It corresponds more closely to the percentage of calories from carbohydrate - regardless of "calories".

Conclusions: Carbohydrate Intake Matters (at least at the low end)

I'm reluctant to extrapolate directly from rats to humans - but these outcomes seem to correspond reasonably well to observed reality in humans.
  • A high-fat diet increased food intake - but the rats didn't get fat, even on over 35% more "calories", unless sugar was added.
  • Just 10% carbohydrate, from sugar, was enough to make a non-fattening zero-carb diet strongly fattening.
  • Even at 10% carbohydrate from sugar, it still took 32% more "calories" to gain the same amount of weight that the rats gained on a 70% carbohydrate diet.
  • It appears that much of the advantage of zero-carb diets is gone at just 10% carbohydrate.
  • The studies I've seen that claim no advantage to varying macronutrient intakes don't reduce carbohydrate to 10% or below. (Or they reduce energy intake to starvation levels, which is a whole another article in itself.)
  • As I've previously warned in this article, almost-ketosis is a bad place to live. You get the pain of trying to adapt to ketosis without ever fully adapting - and, apparently, you also lose most of the associated resistance to weight gain. Most "paleo fails" I see are from hanging around 10% carbs, usually while exercising heavily.
This series will continue! Meanwhile, you can read earlier chapters by going back to Part I, Part II, Part III, Part IV, Part V, or Part VI.

Live in freedom, live in beauty.

JS