Today the Texas House of Representatives Committee on State Affairs will hear HB 708 by Rep. Jessica Farrar. This bill would, among other things, remove from Texas Health & Safety Code a provision that requires patients seeking abortions be informed of the risks of abortion, including “the possibility of increased risk of breast cancer following an induced abortion and the natural protective effect of a completed pregnancy in avoiding breast cancer.”
Abortion is a very emotionally charged topic. On top of that, there is a common meme within the anti-abortion advocacy circles that having an abortion can cause breast cancer. This meme is based largely on early, small studies looking into possible relationships between potential risks and breast cancer.
Because you can find search through the current literature and find many published studies claiming that there is or is not a link between breast cancer and abortion, it is important to understand the scientific process, especially in medical research, to help you understand why some studies are more reliable than other studies.
Studies are expensive, especially studies with stronger methodologies and large sample sizes. In cases such as the cause of breast cancer, we start with a complete lack of knowledge. So researchers will start with weaker studies or studies with smaller sample sizes, in order to produce hypotheses and to eliminate as many hypotheses before spending a lot of money on stronger studies.
The first of two types of studies conducted about the link between abortion and breast cancer, the cheaper and easier type of study, is a retrospective study. This is also called a history study, looking back at the histories of a certain group of people, in this case women with breast cancer, by asking them questions, searching for correlations in exposure to what might be risks. Retrospective studies aren't bad studies, they have their place, but they also have their weaknesses. They are prone to selection bias, because the group is already selected by their affliction, and are even more prone to errors from faulty memory.
Retrospective studies of the link between breast cancer and abortion have had mixed results. Because of this the question has been the subject of longitudinal studies, which are more difficult to conduct, but have stronger methodologies. This type of study takes a group of people and collects data on them in real time as opposed to relying on their memories. This eliminates the weaknesses of selection bias and faulty memory of retrospective studies. Longitudinal studies can have weaknesses as well, most notably attrition bias. To overcome the problem of attrition bias, longitudinal studies conducted on this topic have sample sizes ranging from small to one and a half a million women, while the retrospective studies discussed earlier have had sample sizes of only a couple hundred to a few thousand.
No study is perfect, including the studies conducted on the link between abortion and breast cancer. But in general, studies with more participants are stronger than those with less and studies that minimize weaknesses are stronger than those with weaker methodologies.
Secular Texas has put together a pamphlet that reviews published research on the subject of breast cancer and abortions. We invite you to do your own research as well. What you will find looking at that research is that smaller and weaker studies have mixed results, but are the only studies that find a link between abortion and breast cancer. The studies with stronger methodologies and those with more participants consistently find no link between abortion and breast cancer. This is exactly what researchers would expect to see in the overall literature if there is no link between abortion and breast cancers. In the pamphlet we have divided the studies between meta-analyses, retrospective studies and longitudinal studies.
Secular Texas calls on the Texas legislature to leave the interpretation of science to professionals and to remove laws that attempt to determine what is true through force of law.