Here is the latest post from our guest blogger, Seth Roberts. If you need to get up to speed on Seth’s unorthodox research with weight-loss, mood, and sleep, click here (our N.Y. Times article about him), here (research extras and pix), here (the first round of reader comments), and here, here, and here for his earlier blog postings.
by SETH ROBERTS
Thu., Sept. 15
I told Dubner my students laugh when I say “pimple.”
“They laugh when Levitt says ‘crack cocaine,’” Dubner replied.
My acne began as a teenager. These things go away with time, a dermatologist told me. When, I asked. Everyone is different, she said. I wondered: Doesn’t she know the concept of average? Whatever she prescribed, it didn’t work very well. When I was a graduate student, I saw a different dermatologist. He prescribed both pills (tetracycline, a wide-spectrum antibiotic) and a cream containing benzoyl peroxide. I used them. They too didn’t seem to work very well.
Around that time, I began to do little self-experiments simply to learn how to do experiments. I was following the advice of Paul Halmos, a mathematics professor: The best way to learn is to do, he had written — words that slowly changed my life. The first way they changed my life was when my little experiments expanded to include my acne. It is hard to measure acne, I thought, so I am probably wasting my time. But it was worth a try. I did my best to count the number of new pimples each day. I varied the number of pills I took: Week 1 a certain number per day, Week 2 a different number, Week 3 the first number. The results implied that the pills had no effect. I told my dermatologist this result. “Why did you do that?” he asked, truly puzzled and slightly irritated.
I did more little experiments varying the number of pills per day. The results kept indicating the pills were useless. One day I ran low on pills, so I started to be more careful about using the cream, which I considered useless. My acne suddenly improved two or three days later. It was cause and effect (speaking of delayed causality). You could just look at the time series — one number per day, the count of new pimples — and see this. Just once, if I remember correctly, I stopped using the cream. Two or three days later, my acne got worse. I resumed the cream. Two or three days later, my acne got better. :-)!
That was even better than learning that something was useless. Consistent use of the cream helped a lot. Over the next several years, I only made two further advances. First, I found that a Vitamin B pill helped, probably a multi-B pill. Second, based on the idea of a two- or three-day latency, I discovered that certain foods caused pimples. If my acne suddenly got worse, I tried to remember what I had eaten two or three days earlier. Diet Pepsi and pizza were the main culprits. Taking all this together, I reduced my acne about 90%. Then, as predicted, it faded away.
A few months ago I saw a dermatologist about a different problem. I couldn’t resist: What causes acne? I asked. Well, not diet, he said. I knew this was false. And, really, how could you say such a thing without testing hundreds of different diets — at least? Which no one had. Several years ago I asked my friend Katharine Milton, a Berkeley anthropologist, whether the indigenous people she studies have acne. No, she said, unless they eat foreign food. The anthropologist was right, the doctor was wrong. The situation reminded me of something in an NIH-sponsored online medical encyclopedia: dental malocculsion, it said, “is most often hereditary.” Whereas the work of Robert Corruccini, a professor of anthropology at Southern Illinois University, such as this, had convinced me that dental malocculsion is due to soft food.
It does seem to boil down to laughter, or rather fear of laughter — fear of loss of status. Dermatologists have a fancy word for pimple but they don’t have a fancy word for small. To slowly vary the diet of one patient at a time, year after year, to figure out what dietary elements cause acne is not even close to being sufficiently high-status. It is too risky, not to mention too grubby and too low-tech. It isn’t the sort of thing that produces several publications a year, or a large grant. So a whole class of promising solutions — dietary changes — goes unexplored.