Saturday, September 20, 2014

Does Dean Ornish know how to Google? More on the science of low-carb diets

You've probably heard about the latest study of low-carb diets. It's been covered by every major  media outlet in recent weeks. But in case you haven't heard, the study found that low-carb diets are better for weight loss and improving heart health than low-fat diets.

 It wasn't the first paper to find in favor of an Atkins-like diet and it won't be the last, but because of all the media attention it received, every weight loss guru has weighed in on the study, including Dr. Dean Ornish.

Ornish is a scientist, but it's probably more accurate to call him a vegan apologist, because he's never let the facts get in the way of defending his preferred diet. Though I'm not an expert, I have a functioning brain and access to Google, so it was awfully easy to poke holes in Ornish' review of the Tulane study. His criticisms are neatly arranged into a list, so I'll offer a few comments in the same format.

1. It’s not low carb vs. low fat. It’s both: an optimal diet is low in unhealthful carbs (both sugar and other refined carbohydrates) and low in fat (especially saturated fats and trans fats), as well as in red meat and processed foods.
No, it isn't. Even if a good diet can be low in fat, it's not correct to lump saturated fat together with trans fats and processed foods. We know that people who go on high-fat diets usually lose a lot of weight and see their metabolic health improve dramatically. Study after study after study tells us so. We also have evidence that trans fats and processed foods are unquestionably unhealthy. If you were to reduce your intake of carbs, trans fats and processed foods and up your saturated fat intake, as any good low-carb diet requires, you'd almost certainly lose weight.

2. This study did not distinguish between the types of carbs and fats. Patients in the “low-fat” group of the study increased their consumption of “bad carbs” (sugar and refined carbohydrates) during the study.
Indeed it did distinguish between good and bad carbs and fats. And as the New York Times pointed out, the low-fat diet was in line with the federal government's Dietary Reference Intake recommendations.

3. The “low-fat diet” in this study was not very low in fat. Participants in the “low-fat” group decreased from 35% fat to 30% fat over the course of the study, hardly any change at all. 
Fewer than half your daily calories from fat is a low-fat diet by anybody's standard, except for maybe Ornish'. He says his patients follow a plant-based diet as low as 10% in fat and fare much better than the low-fat group in the Tulane study did, but, I'll reiterate, people who drop the carbs and go on high-fat diets see similar results, and in some studies they even do better than people on the Ornish diet.

4. Perhaps most important, risk factors are not diseases; they are important only to the degree that they affect the underlying disease process. An article published in The New England Journal of Medicine showed what happens inside your arteries on different diets.
Yes, that's true, I guess. But if you can lose weight  and keep it off and you're not showing signs of developing heart disease, it's probably safe to say're not going to develop heart disease. But what of this study in the NEJM? Ornish says it shows that a low-carb diet can clog your arteries. What he doesn't say is that the study involved rats, and the chow they consumed was at best a rough substitute for a low-carb diet actual people would follow. Again, for the third time, we have clinical data showing how people do on low-carb diets. I'll leave it to you to figure out why Ornish cites an animal study instead of one the clinical trials I cited above.

I am not aware of a single study published in a peer-reviewed journal showing that a low-carb diet such as the one described in the Tulane study has stopped or reversed the progression of coronary heart disease. 
But we have research, the Tulane study included, showing that low-carb diets reduce risk factors for heart disease. We also have observational studies showing that people who consume more animal fat are less likely to die of heart disease. I wonder if the two are connected?

Ornish goes on to cite some of his own work showing the benefits of a very low-fat diet. I think he's overselling the benefits of his diet, but I don't fault anyone who wants to try it. Compared to the way most Americans eat, it's far healthier. Nonetheless, it's dishonest of Ornish to trash low-carb diets in the process. I'm forced to conclude that he is ignoring data that doesn't fit his narrative, or he doesn't have access to a search engine. I'm not sure which is more likely.

Wednesday, September 10, 2014

When the evidence changes, so should your opinion

Prominent scientists and science journalists love to proclaim the benefits of science literacy. One point they are particularly fond of is the idea that following the scientific method forces us to change our opinions as new evidence arises. Richard Dawkins comes to mind as the most popular figure to make this point in recent years, but there are others as well.

The unfortunate fact, however, is that many people, scientists included, are fair-weather fans of science; they love science when it affirms their beliefs, but will ignore it when it counters their religious or political views. The evidence of this phenomenon is everywhere.

But there are scientists and educators who buck this disappointing trend. They change their minds when the evidence moves them to, and it's these individuals who really illustrate how our society can benefit from a healthy respect for science. 

Earlier this week, I received an email from Ross Pomeroy, the assistant editor of RealClearScience, one of the best places on the web for science news and commentary, by the way. Ross and I have sparred several times in the last couple of years over the merits of low-carb diets. Following the publication of a big study in the Annals of Internal Medicine which found that low-carb diets are superior to low-fat diets for weight loss, Ross emailed me the following:

The data is swaying to your side, my friend
 ... I'm more inclined to stick to an agnostic attitude on diet. There's evidence that any sort of weight-loss diet can work, so people should do what's best for them. With that said, I don't think the government should be so adamantly pushing their RDIs (roughly 60 - 25 - 15). I think they should definitely come out and say that fat isn't as evil as has been portrayed. 

After I asked him to look at some previous studies that reached the same conclusion as the Annals study he cited, he added:
What I take away from [the studies] is that a LC diet trumps typical calorie restriction in a weight loss setting. After the body has grown obese, what it really needs is something to "shake things up" so to speak 
 It's always validating when someone smarter than you concedes an important point. But beyond my own vanity, Ross' admission is an example of how the process is supposed to work. Researchers make observations, develop a hypothesis to explain their observations and then test it against the evidence. When the hypothesis doesn't fit the evidence, they throw out the hypothesis and come up with a new one, and the world is a better place as a result.

Of course, Ross doesn't have the influence by himself to shift the national conversation about proper nutrition, but there are many people who do--and they're changing their minds, too. The New York TimesTime and Wired have all run stories on the new evidence supporting the efficacy of low-carb diets. Researchers who were openly hostile to the possibility that carbs make people fat are now working with Gary Taubes on a series of studies that may definitively settle the matter.

 This isn't a trivial academic point. Because prominent researchers, writers and even governments (Well done, Sweden) are beginning to accept the benefits of low-carb diets, or at least admit that they're not harmful, average people, when they open their newspapers and turn on the daily news, are going to get accurate information about how they should eat. Provided that these people are willing to act on this new information, we may ultimately see a massive reduction in obesity and the ailments that often accompany it, like heart disease and diabetes. In the long run, that could mean more people living longer, healthier lives.

Think about the implications of such an outcome, and consider the economic costs of obesity while you do. In terms of missed work, higher insurance premiums and lost wages, the costs are significant for everyone. Simply getting accurate information to people is arguably the most important step in eliminating those costs. Add in the fact that being overweight or obese totally sucks, and we have  reason enough to embrace what science has to say about how you make people slim.


Friday, September 5, 2014

There's a gap between e-cigarette science and reporting

Via RealClearScience I learned of a study just published in Environmental Science: Processes & Impacts. The study is a comparison between the emission rates of particulate metals and organic compounds of e-cigarettes and tobacco cigarettes. 

The researchers concluded:

Overall, with the exception of Ni, Zn, and Ag, the consumption of e-cigarettes resulted in a remarkable decrease in secondhand exposure to all metals and organic compounds. Implementing quality control protocols on the manufacture of e-cigarettes would further minimize the emission of metals from these devices and improve their safety and associated health effects.
Given what we already know about e-cigarettes, this paper seems to offer just a little bit more evidence that the smoking alternatives are far safer than tobacco. The original news article on the study, though, reaches a strikingly different conclusion. The article ends with a quote from the director of Smoking Cessation Services at Columbia University Medical Center in New York City:

On the one hand, this study is good news for e-cigarettes; it shows they’re safer,” said Seidman. “On the other hand, it’s bad news for them because it supports the notion of regulating them and they don’t want that.
So secondhand exposure to metals and organic compounds from e-cigarettes is much lower compared to traditional cigarettes, so let's regulate e-cigarettes. How could that possibly follow from the study? At first glance, I wanted to accuse Science Insider of misrepresenting the researchers' conclusion, but the story quotes one of the authors and he seems enthusiastic about banning e-cigarettes in public spaces.

Boston University's Michael Siegel has written about this gap between the data on e-cigarettes and the media coverage that most people see. Not only do we have science journalists misrepresenting the science on this issue, but we have scientists and doctors helping them misrepresent it. What's worse is that e-cigarettes have been subject to critical scrutiny for nearly a decade now, yet this obfuscation continues.

The dishonesty is certainly annoying, but it also means that the anti-smoking lobby is losing the debate--and they know it. Watching their arguments devolve has been a hobby of mine for several years now. When the FDA first weighed in back in 2009, noting that e-cigarette vapor contains  far fewer carcinogens than tobacco smoke, the official anti-smoking line was an argument from ignorance--"we need more data, blah blah blah." Then the clinical evidence started to trickle in, and the tobacco warriors switched to the well worn gateway argument by suggesting that e-cigs lead kids to real cigarettes, which was summarily debunked. Now the time has come, as it does in any public health debate, where the losing side simply begins reinterpreting inconvenient data to suit its agenda, however obvious this ploy may be to anyone who's paying attention.

Ironically, the modern tobacco control movement is behaving exactly as the tobacco industry did many decades ago: they're making stuff up. And a lot of smokers may die as a result.