Is abortion a breast cancer risk?

How’s this for an eye-popping number: Women who have an induced abortion are 20 times more likely to develop breast cancer than those who don’t. The figure appeared in a study of breast cancer risks from Bangladesh that was printed in the Journal of Dhaka Medical College earlier this year. Pro-life media publicized the study last week.

The figure is the largest to be announced among many studies over the past several decades that have suggested a correlation between abortion and breast cancer. In May, a study from India found a six-fold increased risk of breast cancer among post-abortive women.

However, the Indian study was very small—just 188 patients—weakening its overall conclusions. And the Bangladeshi study’s risk figure looks so large because breast cancer is rare in Bangladesh, and most women in the study had few other breast cancer risk factors.

Further complicating the issue is that many other studies dispute the existence of any abortion-breast cancer link. After decades of scientific debate and research about the problem, doctors still disagree about whether abortion contributes to breast cancer. Even many pro-life doctors hesitate to draw firm conclusions.

Breast cancer is the most common cancer afflicting American women, and the second most dangerous. Although early detection and treatment have helped drive down death rates for the past two decades, the disease remains a scourge, responsible for 1 in 36 deaths among women.

The cancer is tied to a variety of risk factors, including diet, age, smoking habits, family history, and whether a woman carries certain genes associated with the disease, such as BRCA1 and BRCA2. According to the government-run National Cancer Institute, a woman who has her first full-term pregnancy before the age of 20 is half as likely to develop breast cancer as a woman who has her first baby after 30. Bearing multiple children and breastfeeding for at least a year also decrease risk.

Many risk factors are tied to the female hormones estrogen and progesterone, which prepare the body for pregnancy and breastfeeding. Oral contraceptives, birth control injections, or hormone therapy after menopause all appear to increase the chances of breast cancer to some extent.

Researchers first addressed the question of abortion and breast cancer in a Japanese study in 1957. The study found that women who had abortions were significantly more likely to develop breast cancer. Many studies in subsequent decades supported that conclusion.

Yet many others found no such link. So many factors influence cancer risk it can be difficult to find a statistically meaningful abortion correlation. Sometimes there seems to be none, and at other times—as in the two recent studies from Asia—the correlation looks unbelievably high.

Indeed, some researchers find the correlation unbelievable. They blame study design for skewing results. Many of the studies suggesting a strong abortion-breast cancer link were based on questionnaires, completed by women who self-reported their abortion history. Critics argue the questionnaire results were influenced by “recall bias”—in this case, a tendency for healthy women with a history of abortions to deny ever having them.

The National Cancer Institute, the American Cancer Society, and the American Congress of Obstetricians and Gynecologists all deny any causal relationship between abortion and breast cancer. They cite prospective studies that tracked the health of post-abortive women over a period of time and found no cancer link.

The gold standard of scientific research, a meta-analysis survey, averages the results of multiple studies on a given topic. In this case, such surveys have come to opposite conclusions. In 1996, Joel Brind, a researcher at Baruch College in New York City, published along with colleagues a meta-analysis concluding abortion increases breast cancer risk among women by about 30 percent, on average.

Brind still stands by the figure. He remains a leading apologist for the abortion-breast cancer link, and is a scientific advisor to the Coalition on Abortion/Breast Cancer and the co-founder of the Breast Cancer Prevention Institute. On Monday, Brind told me researchers who don’t want to show a link publish studies supporting their own view: “They tend to play with the statistics and fudge the data in many different ways.”

As one example, Brind criticizes a major Danish study the American Cancer Society recognizes as “the largest, and probably the most reliable” done on the abortion-breast cancer topic. In it, researchers reviewed the health histories of 1.5 million Danish women born between 1935 and 1978.

“They only used abortions going back to 1973,” Brind said, even though abortion was legal in some circumstances in Denmark beginning in 1939. And although the study tracked abortions and breast cancer up until it concluded in 1992, one-quarter of the women at the end of the study were still under 25, the age group with the lowest breast cancer risk. Brind said these young women were responsible for over 100,000 abortions in the study—but only eight instances of breast cancer.

“Cancer has a latency period,” Brind said. “It’s probably going to [develop] five or 10 years later. It’s not going to be tomorrow.” By excluding very early cases of abortions, and excluding later cases of breast cancer, the study introduced a bias that can make a risk factor disappear, Brind said. (The Denmark study authors have disputed Brind’s criticisms.)

But no one disputes that full-term pregnancy offers a protective effect against breast cancer. By terminating the pregnancy, abortion removes that protection. In the United States, it also limits the protective effect of young motherhood, since abortion is often used to eliminate teenage pregnancy.

If an abortion link exists, how would it work? Brind explains the physiology this way: During pregnancy, estrogen and progesterone cause cancer-vulnerable breast tissue to multiply. This tissue largely becomes resistant to cancer later in the pregnancy, by week 32, when much of it is transformed into milk-producing cells. After a pregnancy reaches this late stage, it confers a long-term protective effect against breast cancer.

But when a pregnancy is unnaturally cut short before 32 weeks, Brind said, the cancer-vulnerable breast tissue never has a chance to develop into the safer, milk-producing tissue. In that way, abortion increases the risk of developing cancer years later. (Brind said an early miscarriage probably wouldn’t have the same effect, because the low hormone levels often associated with miscarriage would produce little breast tissue growth.)

The mechanism might make sense, but even among pro-life doctors, not all are convinced the evidence for the abortion-breast cancer link is conclusive.

The Christian Medical & Dental Associations (CMDA), representing 16,000 healthcare professionals, set up a panel several years ago to review the evidence for and against the abortion-breast cancer link. The organization concluded with this statement, “Studies supporting the different opinions are plagued by imperfect study design. That liability, linked with the potential for author bias, prohibits resolving the question based on the currently available science.”

Dr. Gene Rudd, the senior vice president of CMDA, said he and his colleagues haven’t ruled out the possibility that abortion is a cancer risk. But they don’t want to damage their credibility by overselling evidence that is “concerning, but not convincing.” Rudd said critics of the abortion-breast cancer link do raise some good points, such as the problem of recall bias: “I’m an OB-GYN, and I know that patients often didn’t report abortions to me. They bear guilt over it. There’s some stigma, shame. [They] don’t want their current family to know.”

Rudd has no qualms accusing the scientific community of pro-abortion bias when it comes to reporting on the link, especially among the academic community and science journals in the West: “If they got a good study they would tear it apart. You can always find something you don’t like about a study.” In addition, scientists themselves can present the data in such a way that it shows a link or doesn’t, depending on what side of the abortion debate they fall on.

“Right now, you can sort of throw a dart at the board and land as often on a study that shows a correlation as a study that doesn’t,” Rudd said.

Until research provides a clearer conclusion, CMDA has determined doctors have “an ethical obligation” to inform patients that abortion is a “potential risk” for breast cancer.

That’s a much more careful approach than the one taken by some other medical groups. When influential organizations like the American Congress of Obstetricians and Gynecologists deny a link altogether, U.S. doctors following their recommendations don’t feel any obligation to broach the subject to patients considering abortion.

“I think most doctors don’t counsel on this because they don’t think there’s an issue,” Rudd said. “That’s what they’ve been told.”