Month: October 2009

“Some of Them Will Have the Wrong Answer”

Seth RobertsLeave a comment

In Exploratory Data Analysis, John Tukey tells about visiting a high-school chemistry class. Each student in the class had done an experiment to determine a physical constant. Tukey suggested to the teacher that they gather and plot the results. The teacher didn’t like this idea. Some of the students will have gotten the wrong answer, said the teacher. Tukey didn’t know what to say.

In a previous post, I said there is great stagnation in health care. Obesity and mental illness are the examples most obvious to me, but there are many other problems on which our health care system has made little progress for a long time. (Sure, we should have universal health care but the idea that this will do much about the obesity epidemic, the autoimmune disease epidemic, the autism epidemic, and so on, is absurd. Doctors don’t know how to get people to lose weight. A reasonable health care system would focus on prevention. That is something the current batch of doctors doesn’t know how to do.) I added that a reasonable health policy would empower those who benefit from change.

That’s a difficult thing for people in power to do. Not only does it mean giving up power, it also means giving it to “the wrong people”. The people you like to demonize. People who are . . . not respectable. Not clubbable, John Cheever might say. And, quite apart from that, some of them will have the wrong answer. Tukey’s high-school chemistry class was at a fancy private school, where we might expect such elitist attitudes. But I heard the same thing from colleagues at UC Berkeley when I would suggest giving students much more power to determine what they learned in a psychology class. Some of them will want to learn the wrong things, said my colleagues. I think Tukey was trying to say that the chemistry teacher didn’t understand variability but I think the psychological point of his story is even more interesting.

Effect of Animal Fat on Sleep (more)

Seth Roberts24 Comments

After the striking correlation I described earlier — I ate lot more animal fat than usual and slept longer and had more energy the next day — I started eating much more of what had produced the correlation: pork belly (which is used to make bacon). I couldn’t get uncured pork belly, so I ate bacon. I usually ate it raw. I tried several brands; the only one I liked was from Fatted Calf ($10/pound).

In Beijing I discovered pork belly for sale in every meat department. It is used to make a dish said to be Chairman Mao’s favorite. I bought a soup cooker, an appliance I haven’t seen in America, which made it easy to cook the pork belly. I seemed to sleep better when I had it for lunch.

Finally I did an experiment. I ate pork belly for lunch some days but not others. I ate the pork belly in miso soup, with vegetables. I always ate a whole package of pork belly, which was about 0.7 lb and perhaps 80% fat, 20% meat. On baseline days I ate my usual diet, which was already high-fat by people’s standards. (For example, I ate a lot of whole milk yogurt, a fair amount of nuts, and ordinary amounts of meat.) I tried to alternate baseline and pork-belly days but this wasn’t always possible.
Here are the results on ratings of how rested I felt when I awoke (100% = completely rested = the most rested I have ever felt, 0% = not rested at all).

The lines were fit separately to each set of points (red line to the red points, etc.). The difference is is very consistent (t = 5). Differences in how long I slept were much less clear. I will discuss them in a separate post.

The fascinating thing about this effect isn’t just how clear it is; it’s also how fast it goes on and off (within a day). With most nutrients you’d never see an effect like this. For example, scurvy takes months to develop and a few weeks to recover from. The omega-3 effects I’ve studied have a fast onset but take days to go away.

Sleep is controlled by the brain, of course. The brain is more than half fat, but determinations of how much fat the brain has have measured structural fat. This effect is so fast, both on and especially off, that it must involve circulating fat. Apparently my brain works better when there is a certain amount of animal fat in my blood. This supports Chairman Mao’s idea that pork belly is “brain food” but is a new idea for American intelligentsia. I think the chance that a nutrient that is good for one part of the body is bad for another part is zero — the same as the chance that the electrical appliances you own work best with widely-different currents. The obvious conclusion suggested by this data is that we need plenty of animal fat to be healthy. The only novel element of these lunches was the animal fat. Miso soup with ordinary meat has no effect on my sleep, as far as I know.

I think the science of nutrition proceeds in four steps, repeated over and over for each necessary nutrient: 1. Figure out that we need it. 2. Determine a way to measure how much of it we need. 3. Figure out the optimal amount. 4. Check your answer. With animal fat, conventional nutrition science hasn’t quite reached Step 1. Before this data, I’d say the clearest evidence that we need animal fat is that fat tastes good and long ago we had very little plant fat so it must have been the benefits of animal fat that produced the fat-tastes-good linkage. But conventional nutrition scientists never think this way — never take what we want to eat as meaning anything. And the mere fact that fat tastes good is no help figuring out how much is best.

This data pushes our knowledge toward Step 2. It doesn’t just suggest we need plenty of animal fat for best health, it also makes two methodological points: 1. Animal fat improves brain function. There may be better measures of the improvement than sleep quality. 2. The timing of the improvement — which as far as I know is unprecedented in the study of nutrition — makes it easy to measure.

Yesterday at a Carrefour I watched a pig being cut up. The butcher cut off the skin (with a thick layer of fat) and tossed it into a section of the display of pork for sale. I could buy the part of the pig I valued most for an incredibly low price (about 25 cents/pound). All other pork cost more. That’s how much Chinese shoppers wanted it. No one rushed to buy the newly-cut piece of skin. It reminded me of New York where I tried to buy food past its expiration date, ordinarily considered worthless.

Reduced Diversity of Fecal Bacteria Correlated with Diaper Rash

Seth Roberts2 Comments

In a 2008 study, researchers took fecal samples from 35 babies when they were one week old. They measured the diversity of bacteria in those samples. When the babies were 18 months old, they were divided into two groups: with (n = 15) and without (n = 20) atopic dermatitis, commonly called a rash. Atopic dermatitis is a sign of an oversensitive immune system. Children with this problem are more likely to have allergy problems when they are older. The 2008 study found that the babies with atopic dermatitis had less diverse fecal bacteria.

This is more evidence connecting lack of bacteria with immune problems.

Who is “Totally Healthy”?

Seth Roberts3 Comments

I watched this 60 Minutes piece on swine flu. Of course nothing was said about boosting immunity as a defense. “The best way to reduce your chances of one of those terrible outcomes [hospitalization, death] is to be vaccinated,” said Anne Schuchat, who has a very high position at the Centers for Disease Control and Prevention.

This is just part of a bigger delusion. The story centers on a football player who gets seriously sick after a football game. Schuchat said this:

This is one of the tragic parts of this epidemic. That people who are in the prime of their life, totally healthy, can suddenly become so sick.

Totally healthy. This is the bigger delusion: That the average American who appears healthy is healthy. I believe that practically all Americans have grossly-impaired immunity. Their immune systems work much worse than they could. The poor performance is due to suboptimal sleep and far too little bacteria in their diet. The football player was near death because he had two out-of-control infections. That’s how poorly his immune system was working. And a top CDC official called him “totally healthy”! Apparently she has no idea that people’s immune systems can vary in how well they work. This is even worse than the UCLA medical school prof specializing in infectious disease who also failed to understand this. Schuchat is one of the top public health officials in America! Public health is about prevention. According to Wikipedia, Schuchat “has emphasized prevention of infectious diseases in children.”

The Alternate Universe of Fermented Foods

Seth Roberts3 Comments

In the Afterword to Lolita, Vladimir Nabokov wrote that in his books he tried to create an alternate universe “where art (curiosity, tenderness, kindness, ecstasy) is the norm.” Fermented foods are now a big part of my food world and I remain amazed how different they are from ordinary foods. They are in another universe:

Temperature. To make ordinary food requires high temperatures. You need always be careful that you don’t hurt yourself. Fermented food requires no higher temperature than a hot day.

Deliciousness versus health. With ordinary food there is the tradeoff we are endlessly familiar with: If it tastes good (ice cream, chocolate, cookies) it’s bad for you. If it’s good for you — spinach, carrots, cabbage, brown rice, soy products — it doesn’t taste so great. Anyone who thinks raw food tastes better than cooked food is ignoring history. Whereas fermented food tastes great and is incredibly healthy. (This point has been missed at any number of otherwise great American restaurants, such as Chez Panisse.)

Price. In Berkeley, heirloom tomatoes cost a lot more than ordinary tomatoes. They taste a lot better, too. Perhaps, being organic, they are healthier. The general rule is that better food costs more. An apple costs more than a Coke, etc. Whereas fermented food is often dirt cheap. Kombucha is practically free. For 5 teabags and a cup of sugar, you can make a lot of kombucha. Ordinary milk is cheap but to me at least nutritionally worthless. Whereas yogurt is gold. They cost the same.

Time. Ordinary food takes minutes or no more than an hour or two to make. Fermented food takes somewhere between a day (yogurt) to a month (kombucha) to longer (wine, cheese).

Difficulty. In my experience, it isn’t so easy to prepare a delicious meal if you’re not using fermented food. With fermented food it becomes so much easier. And the result is far healthier, I’m sure.

Need for refrigeration. Fermented food goes bad very slowly at room temperature. Not so ordinary food. I once visited a New York pickle store/factory. No electricity.

You can read a great novel again and again, yes, but not every day. After I read Lolita four or five times, it lost its power over me. But I can happily eat fermented food at every meal, day after day and — judging by other food cultures — year after year.

Too Big to Fail

Seth Roberts10 Comments

An example of “too big to fail” never mentioned in discussions of the financial crisis are big public-works projects: In spite of staggering cost overruns, which occur in practically every project, they are never stopped. The latest example is London’s Crossrail, a new train crossing London. Original estimated cost: 3 billion pounds. Current estimated cost: 16 billion pounds. And construction hasn’t started!

I heard a talk about why this happens. I think the speaker said there was no motivation to be honest. The companies that underbid dishonestly pay no penalty; the politicians that approve their dishonest bids risk nothing. Curiously, in notoriously corrupt China, this sort of thing doesn’t seem to happen (although my Chinese isn’t good enough to be sure). Maybe Dubner and Levitt will write about this in Superduper Freakonomics.

At a talk by Laurie Garrett at the UC Berkeley School of Journalism, I made this point about science journalism: There is no motivation to be honest. The scientists dishonestly inflate the importance of their work, and pay no penalty for doing so; the reporters dutifully write down their lies, and benefit by doing so (because it makes the story seem more important). No, no, this doesn’t happen, said Garrett. Of course it does. The most visible examples are the press releases that accompany the Nobel Prize in Medicine and Physiology, the most prestigious prize in all biology: the whole field should be embarrassed by the claims about the practical importance of teleomere research, which Nicholas Wade dutifully repeated in the New York Times.

There should be some term for these screw-the-public-they’re-too-stupid-to-realize-it situations.

Winter Swimming

Seth Roberts8 Comments

In Jilin Province, where it gets very cold in the winter, the older residents engage in winter swimming. It’s good for their health, they say. Everyone knows this, a friend of mine who grew up there told me. On TV, she once saw an old woman say that she was having heart problems, but once she started winter swimming they got better.

When he was a grad student at Harvard, a friend of mine raised rats to be in learning experiments. He found that if he handled the rats — stressing them, essentially — they grew larger and healthier than unstressed rats.

The cosmic ray effect I mentioned earlier — that trees grow more when there is more cosmic radiation — occurred with older trees but not younger trees.

If you’ve ever designed an experiment, you know that both the treatment and the measurement need to be neither too high nor too low. With the treatment, that’s obvious. I suspect all three of these phenomena are examples of positioning the measurement appropriately. They suggest that everyone needs some sort of stress to be in the best health, but only in certain situations is it easy to see this.

Evidence-Based Medicine

Seth Roberts5 Comments

In the comments, Bruce Charlton writes:

The failure to fund trials is combined with a suffocating dominance of the perspective of self-styled ‘evidence-based medicine’ (EBM) – including the groundless notion that only mega-trails should be taken seriously. . . Since the vast majority of randomized trials are industry funded, EBM has meant that industry has a de facto monopoly on ‘reputable’ therapeutic knowledge.

Delivering us into the hands of Big Pharma was not – of course – intended by the socialistic founders of EBM, but it has happened nonetheless.

This reminds me of something one of my students said. We were discussing male/female differences — in particular, the observation that women are more religious than men. One student said that in her experience, guys were either not religious at all or very religious.

I agree with her. I think this is why EBM has the form it does. Its male founders — not understanding the tendency that my student pointed out — went from one extreme (medical orthodoxy, unrelated to evidence) to another (evidence-based medicine). Reliance on evidence is a good idea, yes, but the founders of EBM couldn’t help making it resemble a religion. You might think that relying on evidence is the opposite of religion but they made the whole thing as religious as possible. EBM became just another way — just another excuse, really — to sneer at people.

Use of Probiotics in Hospitals

Seth Roberts5 Comments

A Canadian company named Bio-K+ makes lactobacilli-based probiotics — mainly a fermented milk drink, like Yakult but with different bacteria — that hospitals can use to reduce antibiotic-related diarrhea (a common side effect of antibiotics) and C. difficile infection, a less common but far more serious side effect. In this 2007 study, the probiotics reduced the rate of diarrhea by half and reduced the rate of C. difficile infection by a factor of 7 (from 7 cases to 1 case).

How the company started. Thanks to Anne Weiss.

Bryan Caplan on Barbara Ehrenreich

Seth Roberts2 Comments

In his blog, Bryan Caplan makes some amusing and reasonable points about Barbara Ehrenreich’s criticism of some happiness research. My eyes widened as I read. This is so much better than what’s usually in the New York Times, The New Yorker, and other publications. It reminded me of Spy, except the level of thought is deeper. It’s as if blogs allow and encourage intelligent people to say what they really think about stuff. Whereas in any mainstream venue there are tremendous constraints.