The cancer study bombshell that wasn't

Were the New York Times and the Washington Post writing about the same New England Journal of Medicine article?

By Arthur Allen
July 14, 2000 11:00PM (UTC)
main article image

Medical quiz du jour: A study published Thursday in the New England Journal of Medicine revealed that A) genes are more important than we thought in causing cancer, or B) genes are less important than we thought in causing cancer.

Correct answer: C) more important if you read The New York Times, less important if you read the Washington Post.


According to the lead of the Post's front-page report on the study, "The vast majority of cancers are caused not by inherited defects in people's genes, as many have come to believe in this age of genetics, but by environmental and behavioral factors."

The Times covered the study with an AP dispatch on page A-21: "Genes may cause more than one-quarter of three major types of cancer, more than previously thought, a group of researchers says."

The reason for this confusion is simple. The New England Journal article, which is based on a large study of twins, essentially tells us what we already knew: that genes and environment both play a role in cancer, just as they play roles in everything else human.


While the article is based on the impressive number-crunching of the medical records of 44,788 pairs of Scandinavian twins, it doesn't contain any major revelations. More than anything, it points to the fact that twin studies, which have long amazed some people and angered others, may be on the verge of outliving their usefulness.

Even though the authors examined the records of a total of 10,803 people with cancer, they got statistically relevant data on only four types of cancer -- breast, colorectal, lung and prostate. Even here the authors acknowledge margins of error of up to 50 percent. The study also lacks data on specific types of environmental exposures that may have contributed to the cancers.

"The study has many strengths but its weaknesses illustrate the difficulties of using data on twins in studies of cancer," Dr. Robert Hoover, director of epidemiology and biostatistics at the National Cancer Institute, writes in an editorial accompanying the study.


"Delineation of the specific environmental and genetic components of the risk of cancer," he adds, "is likely to depend on the emerging new generation of large molecular epidemiological studies, rather than on studies of twins."

The basic idea of twin studies is to gather data on large numbers of fraternal and identical twin pairs and then compare the similarities of the two sets for things like cancer, asthma, schizophrenia and left-handedness.


Identical twins share all of their genes, while fraternal twins are no more identical, genetically, than other siblings, which means they share an average of 50 percent of their DNA.

Twin studies that try to assess the genetic component to things like intelligence and aggression have been a political hot potato, for obvious reasons. But the scientists who conduct them all agree that genes and environment both play a role in behavior, so there isn't really much controversy left.

In the Scandinavian study, the researchers found that 40 of 299 sets of fraternal twins both suffered prostate cancer, while only 20 of the 584 fraternal twins both got the disease.


The Journal's authors, led by Dr. Paul Lichtenstein of the Karolinska Institute in Stockholm, Sweden, cranked those figures through a statistical analysis and came up with the estimate that 42 percent of the risk for prostate cancer comes from heritable factors. Using the same type of analysis, they found that 35 percent of the risk for colorectal cancer and 27 percent of breast cancer risk came from genes.

These numbers mean that the identical twin of someone with prostate, colorectal or breast cancer has an 11, 13 or 18 percent chance, respectively, of getting the same disease. Which means genes are significant to these diseases, but not determinant.

What this tells you is that the Post lead is actually closer to the truth. "The fatalism of the general public about the inevitability of genetic effects," as the study's authors write, "should be easily dispelled" by the data here.


But we didn't really need a twin study to tell us that. Plenty of studies of cancer have shown a significant genetic component.

Still, even if identical twins are no longer needed to prove that genes are involved in various diseases and traits, they can still make a contribution to science -- through the study of their differences.

For example, identical twin sets in which one twin gets breast cancer and the other doesn't offer researchers a chance to figure out the environmental triggers for cancer, by examining the different environmental exposures of the two twins.

"Whether you view genes or environment as more important tends to grow out of your own life experiences," says Nancy L. Segal, a Cal State-Fullerton psychologist who studies twins. "But I think in the future, we're going to see more studies of discordant twins."

Arthur Allen

Arthur Allen writes on health, science and other issues for Salon. He lives in Washington.

MORE FROM Arthur Allen

Related Topics ------------------------------------------

Cancer Environment