Editor’s note. For the latest information about the June 29-July 1 convention, go to nrlconvention.com. The following interview was conducted via email.
Q: What have been the most prominent developments in the ABC link in the year since you last spoke at the National Right to Life Convention?
Dr. Brind: On the scientific front, the most prominent would certainly be the new systematic review and meta-analysis on the ABC link in South Asia (Indian subcontinent) that we at the Breast Cancer Prevention Institute are currently preparing for publication. At this point I can say that the overall risk increase for women who have an induced abortion is far higher than the risk in the West and in China.
I expect that by the time of the NRLC convention, the paper by my colleagues and myself will have been submitted for peer-reviewed publication. As of last year, we had identified 15 primary studies on abortion and South Asia which reported data on the ABC link, and since then, we have found 5 more. So our paper will review 20 studies (unless any more pop up before we submit the paper!).
Q: What about abortion and the Western world? Has any other research been published lately on the ABC link?
Dr. Brind: Another recent development is newly published research from the British actuary Patrick Carroll, who heads the Pension and Population Research Institute in London. Very briefly, Carroll used the long-standing rigid demarcations of social class in the UK to study a worldwide characteristic of breast cancer: The fact is there is a social gradient: wealthier. Better educated women are more susceptible to breast cancer than are less wealthy, less educated women.
This development in breast cancer research provides independent support for the ABC link from a completely different approach than classical epidemiology: comparing long-term trends among large groups of women with different characteristics.
To some extent, there is a correlation between social class and certain known risk factors, such as later age at first childbirth, having fewer children, and breastfeeding them less. However that has failed to explain most of the observed social gradient in breast cancer incidence. What Carroll did is compare the social gradient and breast cancer incidence among the different countries of the UK, i.e., England and Wales, Scotland, and Northern Ireland. A striking finding of his new study is that the social gradient is steepest in England and Wales, where abortion rates are highest, while the gradient is substantially reduced in Scotland, where the abortion rate is lower.
But most striking is that in Northern Ireland, where abortion is still largely illegal and rare, the social gradient almost disappears! So abortion is, as it were, the last risk factor standing to explain, in large part, why breast cancer is a disease of the “higher” classes.
Q: What new or additional weaknesses have sprung up in case against the ABC Link?
Dr. Brind: I’m glad you asked! Now here’s a surprise. As your readers know one of the primary defenses used to dismiss the reality of the ABC link is something called “response bias.” The reason there appears to be more breast cancer among women who have aborted (the argument goes) is because women who do not have breast cancer are more likely than women with breast cancer to deny their abortion history. Or, conversely, those women who have been diagnosed with breast cancer are more likely to come clean about their abortions than are healthy women. Thus it would only appear (falsely) that abortion was associated with breast cancer, due to recall bias.
I have argued for years that there is simply no evidence to support this hypothesis.
But now, I tell you that in studying the South Asian data, I have found evidence of response bias! But it isn’t bias in the direction of overestimating the relative risk of abortion. Rather, it is in the direction of underestimating the risk!
So, without giving away too much from our new study, I have found that in several studies which show little or no risk increase for women who’ve had an induced abortion, there is a tendency for there to be missing data on abortion predominantly among breast cancer patients, rather than healthy women.
It seems that there are several studies in which all of the patients identified as breast cancer patients in a given time period participate in the study, i.e., there is a 100% participation rate among eligible patients. While participation rates are usually high—over 90%–100% is rather unheard of.
So what does this mean? Here’s my hypothesis . Considering how hard high quality medical care is to come by for most women in the relatively non-affluent countries of South Asia, breast cancer patients may be concerned that by not participating it may negatively impact their treatment. Hence, there may be some subtle form of coercion at play. So, when it comes to answering specific questions about sensitive issues such as abortion history, the patients decline to answer these questions, which shows up as missing data in the tables. Meanwhile, healthy women who are asked to participate in a breast cancer study who don’t want to report on such sensitive issues are more likely to simply opt out of participation in the study altogether. So the population of healthy controls is then fleshed out with women who don’t mind disclosing such personal information, and little or no data goes missing.
Q: So what is the bottom line?
Dr. Brind: The bottom line is that since fewer breast cancer patients who had abortions report them, abortion is underreported among patients compared to controls, and the relative risk is underestimated in the study. More on this at the Convention in Milwaukee this summer!
For more information about the annual NRLC convention, which takes place June 29 through July 1 in Milwaukee, Wisconsin, please go to nrlconvention.com.
Go online to register at nrlconvention.com/register.