Two Infinities

Thu, 23 Mar 2006

Dr. Bergren talks to Kiwanis about breast cancer
Dr. Carl Bergren spoke to his Dearborn Kiwanis Club last night about an apect of his surgical practice. I, myself, at my son's recommendation, have been slowly making my way through Atul Gawande's Complications: A Surgeon's Notes on an Imperfect Science. The combination induces a disquiet that, I suppose, accompanies any meditation on our mortality, but there is also an exhilaration at gaining a new perspective on a familiar topic, the practice of medicine.

It is easy to forget how we deal with the world largely through abstractions, and it is easy to forget how difficult and fallible is the process of abstracting from experience. In fact, religion and other human behavior show how abstraction can be distasteful as well as difficult. Surely there is no better example of the jarring confluence of shamanism and science than the medical profession.

Not to put words in his mouth, but Dr. Bergren seemed to feel from his own experience that the increasing incidence of cancers is environmentally related, but at the same time seemed overwhelmed by the difficulty of distinguishing which of the many changes in our environment might be most harmful. He mentioned a population in Japan which exhibited one of the lowest breast cancer rates, but then said when members of that population are removed from their location and diet, they quickly exhibit the same rates of cancer incidence as their new group. (You can imagine how difficult it would be to establish even that one simple statistical fact!) He speculated that the reason might be related to the high consumption of soy products in the Japanese population.

On the other hand, he did mention that there appear to be differences in cancer incidence among racial groups. When questioned further, he said that african-americans tend to exhibit more aggressive forms of breast cancer, but when asked whether that fact might be due to differences in diet and environment, he said only that the issue was a contentious, even political, one. Again, think how difficult a statistical question is being asked. It is hard enough to relate race to genome and near impossible to tease out factors of diet, culture, access to and practice of healthcare.

Another suspect environmental factor Dr. Bergren mentioned several times was growth hormones passed to us through meat. Although he allowed no connection between breast size and cancer rates, he did say that through the years of his practice he felt that there had been an increase women's breast size on average. He didn't seem to feel it was related to obesity, but rather allowed his suspicion to rest on environmental factors like growth hormone. When pressed on this question, shouldn't there be some statistically testable aspect of his hypothesis, he only commented on the difficulty, observing how the similarity of human and synthetic growth hormones fed to livestock made it impossible to relate even the levels of hormone-related chemicals to meat intake.

After considering the difficulties of medicine, Dr. Bergren mentioned the book Doctors Never Lie with which I a not familiar, but he certainly put me in mind of one with which I am: How to Lie with Statistics. And, as I often do, I felt amazed how humans were ever able to abstract anything useful from the morass of physical experience.