Asymmetrical Information

From the desk of Jane Galt:

This Gary Taubes article on

This Gary Taubes article on how we're all getting fat because the government made us that way just won't die.

Taubes asks interesting questions about whether or not fat really gives you heart disease. But then he goes off the deep end and decides that the reason we're all getting fat is that the US government told us to eat simple starches. Carbohydrates, he says, make you fat. Meat and veggies don't.

[Full disclosure: I'm a vegetarian. So I'm not going on Atkins any time soon. But since my diet choices are largely for health reasons, I have a somewhat open mind.]

Taubes doesn't understand the science behind what he's saying. Not that I'm a medical expert (I'm sure I'm about to get fact checked by luminaries like MedPundit and MedRant, so I'll try to keep from straying out too far on a limb). But as it happens, I did some research into the causes of obesity for an article I wrote for Salon, and it's clear to me that Taubes is making some fairly large errors.

First of all, Taubes says that we're fat because most of our diet is made up of grains, etc. -- a high carbo, low fat, moderate protein diet. Absent that, he implies, we wouldn't have gained so much weight since 1980. Umm. . . so how come Africa, where the primary foodstuffs are tubers or grains, isn't bursting at the seams? Or, for that matter, our ancestors, who lived on a diet of up to 80% carbohydrates as late as World War I? Ever eaten on a farm? Sure, you get bacon and eggs for breakfast -- after you wade through a couple of bowls of granny's oatmeal sludge. One or two kinds of bread, canned fruit, and often potatoes as well at every single meal. In fact, the traditional English favorite, Yorkshire Pudding, was the way that farmers wives filled the family up on a wad of bread so that they went easy on the expensive meat. Baked beans, the eternal food of the proletariat, were a meat substitute, loaded with carbohydrates. Yet most farmers weren't fat. And the industrial workers, who were extremely lean, ate even less than the agricultural ones did. The "meat and potatoes man" is an invention of the affluent fifties.

Taubes has engaged in a well-known data mining practice where you carefully choose your start dates and end dates so as to produce a high correlation between two variables that were both spiking during the time period. Another example of this is the way that the New York Times recently mined its data on global warming to produce an enormous warming trend in Alaska, a result the reporter achieved by starting the count on one of the lowest average temperature years on record, and ending it on one of the highest. If he had started the study a year earlier and ended it a year later, says a correspondant in the know, the "warming trend" would have been reduced by more than half.

If Taubes had started in an earlier time period -- say 1950 -- he would have seen no net increase in carbohydrate consumption. But we would have gotten a whole lot fatter. If he'd started in 1900, when most people didn't eat meat every day, we would have seen a net decrease in starch consumption. And the data would still show us blowing up like whales.

Of course, there is a kernel of truth to what he's saying. The science goes like this: simple starches, like those found in white bread and potatoes, turn into sugar in your body much more quickly than fat, protein, or complex carbohydrates wrapped in a package of bulky fiber, like those found in beans, whole grains, or vegetables. While the latter group takes hours to break down into glucose, a candy bar or a potato or a soft, luscious piece of sourdough bread you just pulled out of the bag, start turning into glucose as soon as the enzymes in your saliva hit them.

The result is a spike in your blood sugar. But remember the savannah. On the savannah, blood sugar levels were much steadier, since the body would be breaking down the food you'd consumed over a period of several hours. So while the entire sugar content of that candy bar enters your bloodstream in about 20 minutes, your body thinks you've eaten an enormous amount of berries or cucumbers or bear steaks or what have you, and expects the amount of sugar entering your blood to remain roughly steady for several hours as your meal breaks down. Your endocrine system therefore signals for an enormous amount of insulin to sweep up all that lovely sugar and store it away for a rainy day -- in the form of fat.

The insulin scours out all the sugar in your blood within an hour, leaving you feeling tired -- this is why you get sleepy after lunch, especially a lunch that's heavy and high in starch. Meanwhile, if all you've eaten is a candy bar, your stomach is empty -- and with your endocrine system clamoring for more blood sugar, and no countersignal from a full stomach, your endocrine system tells your brain to get hungry again. So you grab a candy bar to stave off those cravings. . .

So what Taubes has discovered is what the US government has been telling us for years: too much sugar isn't good for you.

But Taubes says the government is the problem: specifically, the anti-fat dictates of the NIH and the FDA's "Food Pyramid". He must not have read the same literature I did. The literature I read specified whole grains, beans, and other complex, vitamin-rich and highly fibrous sources of carbohydrates. It did not say to run out and load up on the fat-free cookies at the snack bar. Taubes glides over this distinction in order to make the case that the government and the medical establishment have been grossly misleading us all these years. Now, I certainly wouldn't put it past them to be wrong, and to cling to their belief long after it was fruitful. But the government did not tell us to up our sugar consumption 30 pounds a year. We did that part ourselves.

For one thing, we stopped looking at calories. Now, I know that there are problems with calories as a measurement. And that two people eating the exact same things won't gain the same amount of weight. But broadly speaking, you will lose weight if you burn off more than you eat, and you will gain weight if you eat more than you burn off. This is the most accurate thing we can say we know about weight loss, because we can stick people in a lab for a month and test how many calories they eat versus how much excercise they do. Long term studies, on the other hand, are notoriously inaccurate. Why? Because when you ask people to self report what they ate, they forget about the three bites of their husband's cheeseburger, the candy at the coworker's desk. . . all the things they wish they hand't eaten. Many of them aim at a sort of wish fulfillment by not telling the researchers. And the bias is all one way, though it's impossible to quantify it -- towards telling researchers that they're eating less of whatever they think they shouldn't be eating.

Any road. The point is that when we got religion on fat grams, we lost religion on not eating anything in excess. My personal trainer used to tell me stories about the neurotic New York women who would swear they were eating healthy -- but would turn out to be consuming a box of snackwells, or eight servings of Tasti-D-Lite (a sort of frozen, artificially sweetened dessert sold here that has the consistency and taste of well-chilled hair cream) a day. These women were only listening to part of the message. Any nutritionist pushing a low-fat diet was also pushing a low sugar, high fiber, heavy on the fruits and veggies and light on the desserts, diet. But these women stopped reading after they saw the words "Fat Free" and were then surprised that they didn't lose weight. So in one sense Taubes has a fair cop -- we listened to the government telling us to eat less fat, and we decided that that meant we should load up on the Snackwells. But that wasn't what the government was saying; it was what we wanted to hear. What we didn't want to hear was that eating healthy meant eating less of stuff we like.

Here's the deal: the stuff that tastes good -- really good -- the stuff that makes you go back to the buffet for seconds and thirds and eighths -- is the stuff that's loaded with

a) Fat
b) Sugar

That's why low-fat foods substitute sugar for the fat (like Snackwells, which have as many or more calories as the cookies they're supposed to replace) and low-sugar foods (like the Atkins diet) substitute fat for sugar. Otherwise no one would buy them. But the culprit isn't carbohydrates, period; it's that while we were cutting our fat intake, our sugar intake pushed our calorie count higher than ever. And I will bet you a million, billion dollars that if the nation went on Atkins, you'd find in ten years that everyone was still getting fat unless they cut the calories they consumed.

Which is actually what Atkins does. Though Taubes pooh-poohs it, the doctors I talked to said that the evidence was quite clear cut that Atkins worked -- where it worked -- by getting people to cut their calorie intake to about 1500 calories a day.

I'm not saying that sugar may not have special bad qualities, just as I think fat probably has its own special bad qualities. But carbs aren't the culprit. Our propensity for consuming massive amounts of processed food that crams as many calories as possible into as little space is.

Posted by Jane Galt at August 21, 2002 10:18 AM | TrackBack | Technorati inbound links