I know why there's such a strong pull towards zealotry in all this diet talk, and it gets me back to sportswriting. I've harped on political coverage before, specifically on the ways in which it falls short of sportswriting.
What seems obvious now is that the lack of good political coverage is what spawns such zealotry in amateur coverage. For eight years, a credulous and uncritical white house coverage pool would simply regurgitate whatever the government told them. The frustration of people who wanted to be informed by rationally critical coverage was palpable. Compare this with major-market sports coverage.
What do you get from sports coverage? You get, for one, easily accessible facts and stats of all manner. Want to know more about last night's game? You can easily find out what to a non-fan would be an astonishing array of facts and stats about seemingly the most trivial details in addition to the major points. It's public, undisputed, easily accessible. One of the nice effects of this is that you can't get away with abject bullshit in the same way you can with political coverage (or health coverage, more on that in a bit). For example, if you fancy yourself a sportswriter or sports expert, you cannot get away with constantly referring to the worst shooter in the NBA as "sweet shooting" unless you're being facetious. You lose credibility because everyone who doesn't already know the guy can't shoot can easily check his accuracy.
The other wonderful thing you get from sports coverage is thorough (arguably insanely thorough) coverage by knowledgeable skeptics. Imagine an NFL general manager says to the press that his priority for the off-season is quarterback. If this were major-market political or health coverage, that would be repeated at face value and that would be the end of it. Because this is sports, this utterance is dissected, dried, dissolved, and discussed ad infinitum. Without having to do any detective work yourself, you, the sports fan, will be told how realistic this is, what free agents might be available, whether they are restricted, which teams would likely sign them, who's available in the draft, what viable trade options are, what your teams salary cap would actually allow, etc. In short order there's a pretty good consensus about whether the GM was simply talking smack, being deliberately evasive, or being sincere.
In a nutshell what you can get from mainstream media sports coverage is a combination of
facts and
critical, informed reporting. This combination -- the accessibility and timeliness of the facts and stats, and events reported on by a critical, informed eye -- can make for a well-informed, critical public. The lack of these qualities in coverage can be striking. Which brings me to health reporting.
Good lord is it bad. What a dire state. (No wonder Ben Goldacre frequently sounds like he's on the brink of taking a hatchet to everyone in sight.) Here we have a (potentially literally) deadly combination: bad science + bad reporting. If the reporting were as thorough, fact-filled, and critical as good sportswriting, the bad science wouldn't be as much of a problem. But bad science reported by the credulous and ignorant is both damaging and enraging (thus the side-effect of zealotry).
Last week a report came out that makes for a perfect case study of both bad science and bad science journalism. Looking at the coverage in the NY Times, USA Today, US News & World, AP, etc., it's clear that the people reporting on it have not read the study. Practices that would get you fired from the sports desk (reporting on a game you didn't watch) are apparently perfectly acceptable when writing about less important matters such as health. So lead researchers Frank Sacks and George Bray publish the study in the New England Journal of Medicine, then summarize it and wax poetic for press consumption, and the press simply regurgitates their talking points. Add a catchy headline that compounds the error, and the public is none the wiser. The public has, in fact, just been made dumber. Reading mainstream health reports is a good way to lose IQ points.
In defense of the public, the facts are neither presented nor easily accessible. The reporting has not been done with an informed, critical eye. If you want to read the study,
Comparison of weight-loss diets with different compositions of fat, protein, and carbohydrates, you go to pubmed and/or find it at the
New England Journal of Medicine. Even after you do that, there's quite a lot you have to know to make sense of it, in particular to see how disingenuous these researchers are being. But the public that relies on these same health reporters for background and context simply won't have any of either.
Here are some sample headlines covering the study:
In 4-diet study, all lost weight if they watched their calories (USA Today)
Study Zeroes In on Calories, Not Diet, for Loss (NY Time)
Want to Lose Weight? Just Eat Less, Diet Study Suggests (U.S. News & World Report)
What's the best Diet? Eating Less Food (TIME)
And some typical ledes:
When all is said and done, it comes down to calories. A landmark study shows that people can lose weight on a variety of diets — including low-fat plans and low-carb ones — as long as they consume fewer calories. (
USA Today)
For people who are trying to lose weight, it does not matter if they are counting carbohydrates, protein or fat. All that matters is that they are counting something. (NYT)
The researchers took a few hundred people, put each person on one of four diets, and tracked them for two years while offering support to help keep to the prescribed diets. These articles faithfully reproduce the researchers' talking points and you walk away thinking they tested, among others, low-carbohydrate diets. Only they didn't. The four target diets included fat/protein/carb ratios of (1) 20/15/65, (2) 20/25/55, (3) 40/15/45, and (4) 40/25/35. Note the first two are considered to be "low fat" while the latter is described as the "low carb" diet. The uncritical reporting fails dramatically here. A carb ratio of 35% is not a low-carb diet. (How would you know that if you were informed only by these journalists? You wouldn't.) Why not try an actual low carb diet (20%, or even better, 10%)? Sacks claimed doing so would not be "realistic". [If physicists did science this way, we wouldn't have quantum mechanics.]
It gets worse. The target diets were aiming for a calorie restriction of 750 kcals/day (i.e. a deficit of 750 calories relative to estimated metabolic rate for each person), but with a lower limit of 1200 calories a day. So even on the lowest-calorie plan, a 35% carb ratio means over 100g of carbs/day -- not low carb. The researchers also asked people to report on what they actually ate, and it turns out that the so-called "low carb" group didn't even average 35%, they averaged over 40% carbs.
What about the results? Basically the message is that all four of their diets were equally useless. Which would have been predictable because these types of calorie restriction diets simply don't work. The group on each diet pretty much lost the same amount of weight the first six months then gained some of it back over the remaining eighteen months. The results were modest: an average 6kg loss after 6 months, but backtracking to an average of 4kg after two years. Remember that the researchers buy into the (repeatedly debunked) caloric balance hypothesis (the sloth/gluttony hypothesis), and that their participants were pushed to follow a 750 calorie daily deficit. So if these guys believe their own theories, the participants should have lost 156 pounds of fat each over the two years. (Sounds ludicrous, but this is exactly the same bullshit that tells you if you switch from one can of coke to one can of diet coke for lunch every day you'll accrue a deficit on the order of 50,000 kcals over a year and thus lose over 14 pounds. And this is the same bullshit they're feeding the press when they say it's all about how much you put in your mouth.) So their theory predicts a loss of 156 pounds over 2 years, versus the observed result of about 9 pounds. Hm. The participants all had starting BMIs of 25-40. For someone 5' 10", this means a weight range of 174-279 lbs. A 9 lb loss shifts their BMI range to 23.7-38.7. An average loss of 6kg over 6 months equates to about half a pound of loss per week. Over two years this drops to less than 1/10th of a pound per week.
This should be enough for them to drop their caloric balance hypothesis. Or at least conclude that their four diets are equally worthless. But instead of headlines saying "High Carb Diets Don't Work" or "Four Tested Diets Ineffective", we get articles saying how vital it is to simply eat less, and how there's no truth to the effectiveness of low-carb diets, written by mainstream reporters credulously quoting the researchers without reading the study or asking difficult questions. It should not be susprising that this sort of thing leads to a bit of fervor amongst the dissenters.