What the Dog Saw: And Other Adventures

division. The argument is similar for endometrial cancer. When a woman is menstruating, the estrogen that flows through her uterus stimulates the growth of the uterine lining, causing a flurry of potentially dangerous cell division. Women who do not menstruate frequently spare the endometrium that risk. Ovarian and endometrial cancer are characteristically modern diseases, consequences, in part, of a century in which women have come to menstruate four hundred times in a lifetime.

In this sense, the Pill really does have a natural effect. By blocking the release of new eggs, the progestin in oral contraceptives reduces the rounds of ovarian cell division. Progestin also counters the surges of estrogen in the endometrium, restraining cell division there. A woman who takes the Pill for ten years cuts her ovarian-cancer risk by around 70 percent and her endometrial-cancer risk by around 60 percent. But here natural means something different from what Rock meant. He assumed that the Pill was natural because it was an unobtrusive variant of the body’s own processes. In fact, as more recent research suggests, the Pill is really only natural in so far as it’s radical – rescuing the ovaries and endometrium from modernity. That Rock insisted on a twenty-eight-day cycle for his pill is evidence of just how deep his misunderstanding was: the real promise of the Pill was not that it could preserve the menstrual rhythms of the twentieth century but that it could disrupt them.

Today, a growing movement of reproductive specialists has begun to campaign loudly against the standard twenty-eight-day pill regimen. The drug company Organon has come out with a new oral contraceptive, called Mircette, that cuts the seven-day placebo interval to two days. Patricia Sulak, a medical researcher at Texas A &M University, has shown that most women can probably stay on the Pill, straight through, for six to twelve weeks before they experience breakthrough bleeding or spotting. More recently, Sulak has documented precisely what the cost of the Pill’s monthly “off” week is. In a paper in the February issue of the journal Obstetrics and Gynecology, she and her colleagues documented something that will come as no surprise to most women on the Pill: during the placebo week, the number of users experiencing pelvic pain, bloating, and swelling more than triples, breast tenderness more than doubles, and headaches increase by almost 50 percent. In other words, some women on the Pill continue to experience the kinds of side effects associated with normal menstruation. Sulak’s paper is a short, dry, academic work, of the sort intended for a narrow professional audience. But it is impossible to read it without being struck by the consequences of John Rock’s desire to please his Church. In the past forty years, millions of women around the world have been given the Pill in such a way as to maximize their pain and suffering. And to what end? To pretend that the Pill was no more than a pharmaceutical version of the rhythm method?

3.

In 1980 and 1981, Malcolm Pike, a medical statistician at the University of Southern California, traveled to Japan for six months to study at the Atomic Bomb Casualties Commission. Pike wasn’t interested in the effects of the bomb. He wanted to examine the medical records that the commission had been painstakingly assembling on the survivors of Hiroshima and Nagasaki. He was investigating a question that would ultimately do as much to complicate our understanding of the Pill as Strassmann’s research would a decade later: why did Japanese women have breast-cancer rates six times lower than American women?

In the late forties, the World Health Organization began to collect and publish comparative health statistics from around the world, and the breast-cancer disparity between Japan and America had come to obsess cancer specialists. The obvious answer – that Japanese women were somehow genetically protected against breast cancer – didn’t make sense, because once Japanese women moved to the United States they began to get breast cancer almost as often as American women did. As a result, many experts at the time assumed that the culprit had to be some unknown toxic chemical or virus unique to the West. Brian Henderson, a colleague of Pike’s at USC and his regular collaborator, says that when he entered the field in 1970, “the whole viral- and chemical-carcinogenesis idea was huge – it dominated the literature.” As he recalls, “Breast cancer fell into this large, unknown box that said it was something to do with the environment – and that word environment meant a lot of different things to a lot of different people. They might be talking about diet or smoking or pesticides.”

Henderson and Pike, however, became fascinated by a number of statistical peculiarities. For one thing, the rate of increase in breast-cancer risk rises sharply throughout women’s thirties and forties and then, at menopause, it starts to slow down. If a cancer is caused by some toxic outside agent, you’d expect that rate to rise steadily with each advancing year, as the number of mutations and genetic mistakes steadily accumulates. Breast cancer, by contrast, looked as if it were being driven by something specific to a woman’s reproductive years. What was more, younger women who had had their ovaries removed had a markedly lower risk of breast cancer; when their bodies weren’t producing estrogen and progestin every month, they got far fewer tumors. Pike and Henderson became convinced that breast cancer was linked to a process of cell division similar to that of ovarian and endometrial cancer. The female breast, after all, is just as sensitive to the level of hormones in a woman’s body as the reproductive system. When the breast is exposed to estrogen, the cells of the terminal-duct lobular unit – where most breast cancer arises – undergo a flurry of division. And during the mid-to-late stage of the menstrual cycle, when the ovaries start producing large amounts of progestin, the pace of cell division in that region doubles.

It made intuitive sense, then, that a woman’s risk of breast cancer would be linked to the amount of estrogen and progestin her breasts have been exposed to during her lifetime. How old a woman is at menarche should make a big difference, because the beginning of puberty results in a hormonal surge through a woman’s body, and the breast cells of an adolescent appear to be highly susceptible to the errors that result in cancer. (For more complicated reasons, bearing children turns out to be protective against breast cancer, perhaps because in the last two trimesters of pregnancy the cells of the breast mature and become much more resistant to mutations.) How old a woman is at menopause should matter, and so should how much estrogen and progestin her ovaries actually produce, and even how much she weighs after menopause, because fat cells turn other hormones into estrogen.

Pike went to Hiroshima to test the cell-division theory. With other researchers at the medical archive, he looked first at the age when Japanese women got their period. A Japanese woman born at the turn of the century had her first period at sixteen and a half. American women born at the same time had their first period at fourteen. That difference alone, by their calculation, was sufficient to explain 40 percent of the gap between American and Japanese breast-cancer rates. “They had collected amazing records from the women of that area,” Pike said. “You could follow precisely the change in age of menarche over the century. You could even see the effects of the Second World War. The age of menarche of Japanese girls went up right at that point because of poor nutrition and other hardships. And then it started to go back down after the war. That’s what convinced me that the data were wonderful.”

Pike, Henderson, and their colleagues then folded in the other risk factors. Age at menopause, age at first pregnancy, and number of children weren’t sufficiently different between the two countries to matter. But weight was. The average post-menopausal Japanese woman weighed a hundred pounds; the average American woman weighed a hundred and forty-five pounds. That fact explained another 25 percent of the difference. Finally, the researchers analyzed blood samples from women in rural Japan and China, and found that their ovaries – possibly because of their extremely low-fat diet – were producing about 75 percent the amount of estrogen that American women were producing. Those three factors, added together, seemed to explain the breast-cancer gap. They also appeared to explain why the rates of breast cancer among Asian women began to increase when they came to America: on an American diet, they started to menstruate earlier, gained more weight, and produced more estrogen. The talk of chemicals and toxins and power lines and smog was set aside. “When people say that what we understand about breast cancer explains only a small amount of the problem, that it is somehow a mystery, it’s absolute nonsense,” Pike says flatly. He is a South African in his sixties, with graying hair and a salt-and-pepper beard. Along with Henderson, he is an eminent figure in cancer research, but no one would ever accuse him of being tentative in his pronouncements. “We understand breast cancer extraordinarily well. We understand it as well as we understand cigarettes and lung cancer.”

What Pike discovered in Japan led him to think about the Pill, because a tablet that suppressed ovulation – and the monthly tides of estrogen and progestin that come with it – obviously had the potential to be a powerful anti-breast-cancer drug. But the breast was a little different from the reproductive organs. Progestin prevented ovarian cancer because it suppressed ovulation. It was good for preventing endometrial cancer because it