I was excited to see an article, “A Look at the Low-Carbohydrate Diet,” in the December 3, 2009, New England Journal of Medicine. I was quickly disappointed.
Expecting a scholarly review of low-carb eating in humans, I found an exposition of a diet study in mice. And not just your garden-variety mice. These were a lab strain deficient in apolipoprotein E, which makes them particularly susceptible to atherosclerosis when fed a “Western” high-fat, moderate-protein, moderate-carbohydrate diet instead of standard lab chow.
Click on the HeartWire reference below for a discussion of the original mouse research. I wrote a short post about it in August, 2009.
The article author, Dr. Steven R. Smith, states the usual concern that high-fat (especially saturated fat), high-protein, low-carb diets may cause cardiovascular disease such as atherosclerosis (hardening of the arteries). He doesn’t mention the scientific evidence showing little or no role of total and saturated fat in cardiovascular disease.
I give credit to him for mentioning that high-fat low-carb diets area associated with improvement in several cardiovascular risk factors such as HDL cholesterol and blood pressure. He thought they also improve ( lower) LDL cholesterol levels—not something I’ve been impressed with. He didn’t mention the lowering of triglycerides so often seen.
Dr. Smith explains that, compared with controls, mice eating the Western high-fat low-carb diet demonstrated progression of atherosclerosis, perhaps mediated by elevated nonesterified fatty acids and low numbers of endothelial progenitor cells. These are not yet considered classic cardiovascular risk factors in humans.
To quote Dr. Smith, his main point is that . . .
The work of Foo et al suggests that the [high-fat low-carb] diet might increase the risk of cardiovascular disease through mechanisms that have nothing to do with these “usual suspects” [e.g., LDL and HDL cholesterol, blood pressure, C-reactive protein] and so provides a note of caution against reliance on the traditional cardiovascular risk factors as a gauge of safety.
He rightfully calls for investigation of these issues in humans, but . . .
In the meantime, the ageless advice applies to the consumer of the [high-fat low-carb] diet and other fad diets: caveat emptor.
Take Home Points
I agree that human studies are needed.
As the evidence in favor of the safety and efficacy of high-fat low-carb diets increases, the reigning medical establishment is looking for new ways to discredit them. This attempt is pathetic.
Unfortunately, the typical physician reading NEJM will skim this article and conclude, “Yeah, I was right—the Atkins diet causes heart disease. Low-fat high-carb is still the best.”
If you have beloved pet mice that are deficient in apolipoprotein E, don’t feed them a high-fat low-carb diet.
References:
Smith, Steven R. A Look at the Low-Carbohydrate Diet. New England Journal of Medicine, 361 (2009): 2,286-2,288. [This may cost you $10 USD.]
Foo, S.Y., et al. Vascular effects of a low-carbohydrate high-protein diet. Proceedings of the National Academy of Sciences of the United States of America, 106 (2009): 15418-15423. doi: 10.1073/pnas.0970995106 [This may cost you $10 USD.]
Busko, Marlene. Atherosclerosis heightened in mice fed low-carb, high-protein diet. HeartWire, August 26, 2009. [Free]
Diabetic Mediterranean Diet 
That study is a joke. Having worked with ApoE and LDLr mice for over a decade, where do I even begin? This study was clearly not done by anyone with a physiology background nor a good grasp of disease etiology and it amazes me that it made to NEJM for a review article.
1 – ApoE are not a good model for obesity. They don’t get that obese, nor do they reliably develop insulin resistance.
2 – ApoE mice are impaired in the clearance of VLDL/chylomicron remnant particles. Why in the world would one choose them to look at the effects of dietary lipids when altering macronutrient levels? I don’t have access to PNAS anymore. Did they even bother to measure food consumption or look at the lipoprotein profile?
3 – ApoE plaques look nothing like a human plaque at mid/later stage. No plaque rupture and no thrombotic events in these guys despite the similarities of initial plaque formation.
4 – Overall, they are very poor models for human dyslipidemia (all VLDL/little LDL, no CETP, very high LDL clearance, etc). Case in point, statins do NOT work in this model but ezetimibe DOES. Kinda the opposite in humans. I could continue but that last point should drive it home.
Thanks for the input, Isaac.
-Steve
To expand a bit, the whole point of the low carb diet is to lower A1C levels which are known to be a strong risk factor for mortality, not to necessarily improve lipoprotein profiles (though it may have some effect on that, whether beneficial or not is still up for debate). It’s to reduce heart disease by reducing diabetes. The ApoE model doesn’t even begin to address that question because they don’t even get hyperinsulinemic, much less develop elevated A1C levels. It looks at high protein and high fat in the context of lipid levels when they’re at superphysiological levels, cholesterol especially. It completely removes the diabetes aspect and looks at lipids in a highly artificial model of dubious relation to the human condition. To me, it’s asking a question using the wrong tool.
Had you seen this study?
http://www.theheart.org/article/1033837.do
I’m not familiar with the aortic agumentation index so I can’t really judge the quality of the outcome.
I hadn’t heard of aortic-augmentation index before.
For those who don’t click on the link: a UK study seemed to show increased arterial stiffness in overweight people on a low-carb (20% carb) weight-loss diet compared to a low-fat (20% fat) weight-loss diet.
At least they weren’t studying mice.
The issue bears watching.