Last week I told you about a study that got a ton of press in the media, a study that concluded that animal-based low-carb diets were associated with greater rates of death over time than vegetable-based low-carb diets.
In the intervening week, I’ve had a chance to interview Dr. Teresa Fung, the lead author of the study, and I’ve also had a chance to read some of the scathing reviews of the study posted on the blogosphere (linked below).
Let me start by saying that despite what many of my highly esteemed colleagues in the low-carb community have been saying, these researchers are very far from idiots. Dr. Fung is a delightful, forthcoming and exceptionally bright researcher who has the best of intentions and certainly isn’t the evil Satan many of my colleagues have made her out to be. She may have begun the research with some widely shared misconceptions about what low-carb diets really are, but that doesn’t make her a bad- nor a stupid- person.
That said, I believe the conclusions of this study are deeply flawed, and the reporting on it—especially by people like Dean Ornish—has been disgracefully inaccurate. (Ornish titled his article “Atkins Diet Increases All-Cause Mortality”– the Atkins Diet wasn’t even mentioned in the study.)
Because there have been some very careful and detailed discussions of the actual data elsewhere (see links below), I’m going to summarize some of the most important points here and suggest that you read the articles referenced below if you want a more detailed analysis.
1) This was an observational study, not an experiment. The researchers examined food questionnaires that had been collected over the course of a couple of decades and then put people into categories (low-carb/animal sources, low-carb/vegetable sources, high-carb) depending on what their food questionnaires showed they were eating.
2) The researchers “ranked” the subjects according to the percentage of calories the subjects were getting from protein, fat and carbohydrate. Those eating the least amount of carbs in comparison to the rest of the population were put in the low-carb group, and then further divided into ten “deciles”. The “low-carbers” eating the highest amount of carbs were in the first decile, those eating the lowest amount of carbs in the tenth decile.
3) Those who were considered to have a “low-carb” diet consumed between 35% carbs (the tenth decile) to 60% carbs (the first decile). Many critics have correctly pointed out that this range can hardly be considered a true “low-carb” diet. Barry Sears, whose program is around 40% carbs (right near the “bottom” end of the carb consumption in this study) has frequently pointed out that his diet is not low-carb. So it’s hard to see how a diet of 35% carbs qualifies as low-carb, let alone a diet of 60% carbs!
There’s something else, and it’s important. This study- and many others like it- talk about protein, carbs and fat in terms of percentage of total calories. This is a common shorthand for characterizing diets, but it’s not adequate and to show you why it helps to use an example from your household budget.
Let’s say you’re spending 30% of your budget on entertainment. That might sound like a lot- but what if your whole budget is 10 bucks? Then the total amount of entertainment money is $3, not enough to get into a movie even at senior rates during the afternoon. This “total” amount- three bucks- is hidden in the data if you just look at percentages.
Applying this to protein, fat and carbs, if you’re eating 1500 calories a day and 20% comes from protein, you’re only eating 75 grams of protein, hardly an “excessive” amount. If you’re eating 3000 calories, that same 20% adds up to 150 grams of protein. And if you’re a bodybuilder eating 5000 calories, 20% of which come from protein, you’re consuming a whopping 250 grams of protein. All programs have the same percentage of protein (20%), but they are hardly identical.
In this study, calories reported overall for the “low-carb” eaters ranged from 1641 to 2034. (I don’t know about you, but based on what I know about how people eat and what we all know about the obesity epidemic, these numbers sound awfully low, but let’s go with it.) The men in the 10th decile of low-carb/ animal sources group reported eating an average of just 1881 calories (!) and had an average BMI of 26.5– only slightly overweight. (Right, and that actress on the cover of People claims she never had Botox.)
These same men reported eating about 18% of their calories from animal protein. That means they were consuming 84 grams of animal protein a day. It’s very hard to see how that constitutes a diet high in animal protein. So if these folks had a higher risk of dying, blaming it on animal protein seems to me quite a stretch.
And speaking of animal protein, here are a few points that Dr. Fung herself made to me when we discussed the study:
1) Animal and fish sources were lumped together in this study. I think that’s interesting because no one- let alone the researchers- jumped to the conclusion that a high fish diet would increase mortality, even though fish was put in the “animal” category.
2) No distinction was made for how the “animal” products were prepared. Seven servings a week of fried fish “nuggets” and some charbroiled steaks with few vegetables and fruits would certainly put someone in one of the higher deciles of “low-carb”, but is it the “low-carbiness” of the diet that is a problem or is the fact that they’re eating fried food and bar-b-que? “It’s known that when meat is cooked in high heat and the protein is burned it creates carcinogenic (cancerous) compounds”, Dr. Fung pointed out.
3) Finally, no distinction was made between factory farmed meat and grass-fed organic meat. In fairness to these researchers, no study I’ve ever seen has looked at that variable, but that doesn’t mean it’s not important. People eating grass-fed burgers are lumped together in the “animal” category together with people eating high-sodium baloney, blackened steaks from factory farmed hormone-laced cows, and fried fish. It doesn’t pass the “smell” test that all these foods have the same health impact.
Much has been made of the fact that the people put in the “low-carb/ animal” group were considerably unhealthier—they smoked more and tended to be overweight more. Researchers try to “correct” for the influence of these variables statistically, but the corrections are usually imperfect.
When people are smoking and overweight they tend to do other unhealthy things that the researchers might miss when “correcting” for the known variables like smoking. Perhaps these folks sleep less, get less sun, have more stress, or take more meds, prescription and otherwise. We just don’t know.
On this subject, Denise Minger, one of the most vocal critics of the study, had this to say:
“Folks in the Animal Group were more likely to smoke and had higher BMIs than adherents of the Vegetable Group. Along with influencing mortality outcomes, this suggests the Animal Food group, in the aggregate, may have been somewhat less health-conscious than the dieters lumped into the vegetable category. And that’s the type of thing that has repercussions for other diet and lifestyle choices that weren’t measured in the study.”
Minger goes on to say that what this study really measured was a Standard American Diet (highest Animal Group decile) and a slightly-less Standard American Diet (highest vegetable group decile). And I’m afraid I have to agree with her conclusion, which is pretty hard to improve on:
“Both (groups) ate sucky diets, but the latter had slightly less suckage. You can bet the farm that neither was anything close to “low carb.” And if you have two farms, you can bet the other one that neither diet group was anything near plant-based, so I’m not sure the vegan crowd has much to gloat about here.”
For those wanting more detailed critiques of the study, I recommend three articles: Denise Minger’s “Are Low-Carb Meat Eaters in Trouble?”, Tom Naughton’s “The ‘Atkins’ Study (ahem, ahem) According to Ornish”, and the excellent (if somewhat dense and detailed) analysis by Chris Masterjohn.
As for me, I’m sticking with a controlled carb diet of around 100-120 grams a day of carbs mostly from vegetables, fruits, and the occasional oatmeal or slice of Ezekiel bread. I get my protein from grass-fed meat only, plus wild salmon and tuna, free-range chicken, whole eggs and whey protein.
As for fat? As long as it’s not trans-fat, I worry about it not one whit.