But wait, it's not just a low-carb diet, it's an especially high-protein diet and to be fair to the researchers that seems to have been their intention. In the first place to change only one variable - so they held the fat content steady with respect to the western diet and cut the carbs. To keep the diet iso-caloric (to avoid the confounding effect of energy intake differences) they had to increase another macronutrient, so they chose protein. A second reason for this might be that ,in general, low-carb diets for humans (which should really be high fat diets (43% No?! really high-fat - try 68%)) are often camoflauged as high-protein diets to smoothe the sensibilities of the fat-phobes. And to be fair again, the authors do pass comment that the results point out effects of macronutrients other than fat.
So, first reactions:
- the diet is really, really high in protein - 45% - this could be directly bad for the mice, for these particular mice or for any mice. For humans, it's well-known that there is a physiological limit to protein intake - 30-40% is the often quoted figure. This quote is from Loren Cordain's paleo diet faq.
It is physiologically impossible to gain weight when lean protein is the only food consumed because of the body's limited ability to break down protein and excrete the by-product of protein metabolism (urea). This limit is called the physiological protein ceiling and varies between 30-40% of the normal caloric intake in most people, assuming they are consuming their usual (eucaloric) energy intake. Continued consumption of lean protein at or above the physiological protein ceiling without added fat or carbohydrate will elicit symptoms of so-called "rabbit starvation," a malady eliciting lethargy, diarrhea, weight loss, electrolyte imbalances, and eventual death. Hence, all people will lose body weight if limited to consumption of lean protein.
- Most studies of the danger of high protein intake focus on renal consequences and one study on rats found no ill consequences of a 50% protein intake (however, the study was really short-term). It's extremely difficult to establish what other issues there might be with a very high protein intake.
- The protein given to the mice was casein. While not wishing to add to the hysteria over milk and heart disease (and being a cheese lover), it has to be pointed out that milk protein has some form (admittedly in the shape of weak epidemiological evidence) in the causes of CHD stakes. There is an extensive discussion in the comments at Whole Health Source here. And despite all the folklore surrounding mice and cheese, mice aren't really cheese-eaters.
To see this consider the following example: a 2000 calorie western diet with 42% carb/43% fat/15% protein has 210g carbs, 95g fats and 33g protein whereas a 2000 calorie 'low-carb' diet with 12%C/43%F/45%P has 60 g carbs, 95g fat and 100g protein. Let's say 15% protein or 33g matches the body's protein needs, that means in the second diet there are 67g extra. While this doesn't mean that 67g of carbs will be generated, it does mean that the diet does have greater parity in the quantity of carbs and fats presenting for energy provision, which isn't a good thing physiologically.
Update: this probably isn't the reason the mice on the 'low-carb' diet did worse than the western diet mice. Apart from the simple fact that apoE -/- mice are doomed by their genetic oddity, there seems to be an extra issue with the high protein -specifically the casein which leads to kidney damage and a specific effect on those cells which can regenerate blood vessel damage - see Peter's updates here.