Abortion and Breast Cancer

New York Times
October 12, 1996

Analysis of Studies Suggests Abortion Can Raise Breast Cancer Risk
by JANE E. BRODY

A combined analysis of 23 studies has suggested that induced abortion can slightly raise a woman's later risk of developing breast cancer. The reported increase in risk of about 30 percent is considerably less than might result, say, from having a family history of breast cancer or delaying childbirth until after age 30.

The analysis, published Friday in a British journal, The Journal of Epidemiology and Community Health, sought to bolster the statistical value of dozens of individual studies, most of which had found a weak relationship between abortion and breast cancer.

The research team, headed by Dr. Joel Brind, a biochemist at Baruch College in New York, reported that it had included all studies, good and bad, in which induced abortion had been examined as a risk factor for breast cancer so as not to be accused of selecting those that favored a particular conclusion.

But some experts questioned the significance of the finding and the statistical validity of combining the results of studies that relied on reports by patients who developed breast cancer that they had had abortions.

In the introduction to the study, Brind, who has spoken against abortion and has written about the relationship between abortion and breast cancer in the National Right to Life newspaper, asserted that the medical community had failed to warn abortion clients about this risk and had generally ignored it in medical articles about the risk factors.

Dr. Michael Burnhill, vice president for medical affairs at the Planned Parenthood Federation of America, said the new study, "which lumps together studies done in different countries over the last 40 years, shows a very small increase in relative risk" of breast cancer following abortion.

He said his organization would wait for the study to be "analyzed by dispassionate experts at the National Cancer Institute" before making a policy decision as to whether women seeking an abortion should be counseled about this risk.

Dr. Carolyn Westoff, an obstetrician-gynecologist and epidemiologist at the Columbia University School of Public Health and chairwoman of the medical board at Planned Parenthood of New York City, said, "As a clinician, I haven't felt there was enough evidence of a relationship to bring this up with my own patients." She said abortion counseling generally focuses on the pregnancy itself, the medical issues involved in the procedure and its effects on future fertility.

Another researcher familiar with the Brind analysis challenged its validity and significance and questioned its assumptions. Dr. Lynn Rosenberg, an epidemiologist at Boston University School of Medicine, said: "A relative risk of 1.3 (that is, a 30 percent increase in risk) is in epidemiological terms virtually indistinguishable from a risk of 1.0 (that is, no increase in risk). Even if the finding is true, it has no meaning for the individual woman because the change in risk is so minuscule as to be not worth considering."

Cigarette smoking, by contrast, increases the risk of developing lung cancer by 200 to 800 percent. With regard to breast cancer, the increase in risk is very much higher than 30 percent among women who have a strong family history of the disease, women who have never had children and women who have their first child after age 30.

Rosenberg also questioned the significance of the new finding because studies of women who had more than one abortion showed no further increase in their risk of developing breast cancer. If abortion can result in breast cancer, she said, there should be an obvious dose-response relationship. That is, the higher a woman's exposure to the supposed cause, the greater her risk should be.

Rosenberg also challenged the researchers' use of the statistical technique called meta-analysis on the kinds of observational studies that have suggested there is a link between abortion and breast cancer. Meta-analysis is primarily used to increase the chances of establishing a statistically significant result by combining a large number of experimental studies, some of whose findings may not have been significant because of their small size.

The statistical technique also seeks to cancel out biases that may be inadvertently introduced in individual studies. But in observational studies "the biases are all likely to be in the same direction and so do not cancel out one another," Rosenberg said.

A potential bias is the reluctance of many women to admit having had an abortion. Researchers suspect that women who develop breast cancer may be more willing to reveal potentially embarrassing facts about their past, including having once had an abortion.

Dr. Vernon Chinchilli, a biostatistician at Hershey Medical Center who provided the statistical expertise for the new report, said in an interview, "I think the increase in risk is real because 19 of the 23 studies showed it, although it is still possible it's due to an accumulation of bias."

Chinchilli said that "even though a 30 percent increase in risk is not a large increase, it is of concern because exposure is so high." He said that "800,000 women undergo abortions each year, so a small increase in risk can lead to a reasonable number of breast cancer cases." The researchers projected that at the current rate of induced abortions, 9,000 to 40,000 additional breast cancer cases a year may result decades later.

Chinchilli said he thought women seeking abortions should be made aware of this small increase in risk, although he added, "It should be up to the individual woman to decide how to react."

In their analysis, the team sought to isolate the effect of abortion on breast cancer risk apart from other factors, such as whether the women involved had children and at what age and whether childbirth had occurred before or after the abortion.

It has long been known that giving birth to a child, especially before age 30, helps to protect against breast cancer, presumably because of hormonal changes during the later third of pregnancy.

Animal studies have suggested that estrogen secreted early in pregnancy stimulates the multiplication of immature cells in the breast and that these cells do not mature fully until the end of pregnancy. Theoretically, then, when a woman's pregnancy is cut short through induced abortion, a lot of immature cells remain in the breast that are vulnerable to cancer-causing influences.

--------------------------------------------------------------------------------

. . . .