Science and abortion
Okay. This one, I gotta say something about. It's about abortion, breast cancer, and science. Abortion and breast cancer are the proximate subject, but science is the ultimate one.
http://www.alternet.org/reproductivejustice/95249/?page=1
http://www.alternet.org/reproductivejustice/95249/?page=2
http://www.alternet.org/reproductivejustice/95249/?page=3
http://news.bbc.co.uk/1/hi/health/1692144.stm
http://www.abortionbreastcancer.com/abc.html
cancer.org
http://www.news.uiuc.edu/NEWS/06/1201abortion.html
Now.
One study is not sufficient evidence.
Two studies are not sufficient evidence.
Three studies are not sufficient evidence.
Especially when they disagree.
Scientists err. All the time, they err. We cannot tell when they err, except by studying the study. Even then, we cannot really tell when they err, but we can (often) tell that they may have erred.
The fact that bias might be present does not mean that it is present. However, if a study could have bias, which way might it be biased? How can we tell?
Here's the bias identified in some studies:
Studies have shown that healthy women are less likely to report that they have had induced abortions. In contrast, women with breast cancer are more likely to accurately report their reproductive histories because they are searching their memories for anything that may have contributed to their disease.
Does that mean that this particular bias is present in those particular studies?
Actually, probably.
How much was this bias? Can we tell?
Okay, let's talk about the studies mentioned at cancer.org. These are large prospective studies. That is, they choose cohorts of women and review their entire history after a large period of time; the women's entire health-care records are available for review. In other words, there is no chance for recall bias. Other forms of bias, like reviewer bias - let's not go there right now.
And, there are several of these studies, including a study of studies.
The conclusion is that there is no link between abortion (of any sort) and breast cancer.
Do I trust this conclusion? Tentatively. Why not absolutely?
Because scientists - even large groups of them - err all the time. In the 1960's and 1970's, the concensus of opinion about plate tektonics was that it was hokum. Nothing could be published in peer-reviewed journals about it. Until somebody got some editor to show the evidence; then it was like the dam burst - there was all kinds of evidence. BUT, for a long period of time, plate tektonics was "nonscientific" and "unsupported by peer review" and thus could not be talked about. Now, of course, it's accepted as valid.
Let me discuss reviewer bias. When a bunch of experts are invited to discuss a particular topic, who gets invited to the panel? The ones the empaneling authority agrees with? How can that be unbiased? So, even though a large number of scientists were around a table, that does not mean that they were unbiased. It also does not mean they were biased; we simply do not know.
Big tobacco was an expert at creating concensus, and we we know exactly how biased their panels were. We also know their motive: money.
However - what reason could the National Cancer Institute (NCI) have for promoting such a bias, if one were to exist? Is it likely that the NCI would empanel a biased set of reviewers? What would be their motive?
I don't like conspiracy theories. And I'm not going to advance one here. I hope that someone would blow the whistle on that kind of thing.
I haven't looked at the studies. I haven't looked at the questions the researchers were asking. So I don't really know whether or not to trust them. But that many studies came to the same statistical conclusion indicates that I should, at least partly. (Many reviewers, many viewpoints, same conclusion... okay; we'll go with it for now)
However, statistical studies control for other causes; that is, if one activity that's known to be a risk factor is factored out, other things that cause that activity are hidden from view. Therefore, in order to know what's going on, we need to look at a different question.
This site raises questions about these studies, by providing biological evidence for the site's position. This site is clearly biased. Does that mean they are wrong? I'm not so sure. You see, we know that childlessness increases the risk of breast cancer. It is not the abortion that's the risk factor, but childlessness that increases the risk of cancer (one of four major factors). When abortion is the cause of childlessness, well, you'd think that'd be a conclusion. But, actually, that's not the case; people who've had children also have had abortions.
Here is the press release about a study that links abortion and breast cancer.
Hm. Okay. We have conflicting studies and conflicting evidence. Who do we believe? The biggest? The mostest? Maybe not. We can ask a different question, though.
The ACS reports a link between reduced risk of breast cancer and childbearing at an early age.
After looking at all of this mess, I must conclude that many scientists and doctors are not looking at the right things or asking the right questions. My question would not be, "Does abortion cause breast cancer?" but rather, "What are the risk factors associated with breast cancer?" Just as I should ask, "What are the risk factors associated with lung cancer?"
Causation is hard to identify; risk factors are easier.
Abortion probably does not cause breast cancer (in the face of millions of cases reviewed, I will not claim it does). However, people should be aware of what the risk factors are. There are four primary risk factors associated with breast cancer:
1) Childlessness;
2) Small family size;
3) Little or no breastfeeding; and
4) Having a late first full term pregnancy.
What to conclude about abortion? That it was not statistically linked to breast cancer means that abortion is evenly distributed among the population of women of childbearing age.
To use another example: spending money on cigarettes is connected to lung cancer, but does not cause it. It's not the spending that's the problem; it's the smoking. In a similar way, abortion is connected to breast cancer, but does not cause it.
To announce that abortion is not linked to breast cancer is to tell a truth - but not the vital truth concerning the risk factors of breast cancer. In a sense, to vaunt the nonconnection is to deceive the listener or reader. If small family size is a risk factor for breast cancer, and if choosing abortion is the cause of small family size - you can see the connection, even though abortion is not a risk factor itself. That does not mean that choosing not to have a family is wrong. It just means there are risks associated with that decision. But to have a larger family can also mean higher risk. Balancing risks is part of being human. There are risks associated with all parts of life, even in seemingly non-risky behaviors.
The choice of whether to abort or not is a decision best reserved for the woman who is pregnant. That woman should be well-informed of the risks, benefits, and results of that choice, whichever way she chooses, and she should not be pressured one way or the other by the health-care professional who is offering the choice (that would be unethical).
I hope that made sense.
http://www.alternet.org/reproductivejustice/95249/?page=1
http://www.alternet.org/reproductivejustice/95249/?page=2
http://www.alternet.org/reproductivejustice/95249/?page=3
http://news.bbc.co.uk/1/hi/health/1692144.stm
http://www.abortionbreastcancer.com/abc.html
cancer.org
http://www.news.uiuc.edu/NEWS/06/1201abortion.html
Now.
One study is not sufficient evidence.
Two studies are not sufficient evidence.
Three studies are not sufficient evidence.
Especially when they disagree.
Scientists err. All the time, they err. We cannot tell when they err, except by studying the study. Even then, we cannot really tell when they err, but we can (often) tell that they may have erred.
The fact that bias might be present does not mean that it is present. However, if a study could have bias, which way might it be biased? How can we tell?
Here's the bias identified in some studies:
Studies have shown that healthy women are less likely to report that they have had induced abortions. In contrast, women with breast cancer are more likely to accurately report their reproductive histories because they are searching their memories for anything that may have contributed to their disease.
Does that mean that this particular bias is present in those particular studies?
Actually, probably.
How much was this bias? Can we tell?
Okay, let's talk about the studies mentioned at cancer.org. These are large prospective studies. That is, they choose cohorts of women and review their entire history after a large period of time; the women's entire health-care records are available for review. In other words, there is no chance for recall bias. Other forms of bias, like reviewer bias - let's not go there right now.
And, there are several of these studies, including a study of studies.
The conclusion is that there is no link between abortion (of any sort) and breast cancer.
Do I trust this conclusion? Tentatively. Why not absolutely?
Because scientists - even large groups of them - err all the time. In the 1960's and 1970's, the concensus of opinion about plate tektonics was that it was hokum. Nothing could be published in peer-reviewed journals about it. Until somebody got some editor to show the evidence; then it was like the dam burst - there was all kinds of evidence. BUT, for a long period of time, plate tektonics was "nonscientific" and "unsupported by peer review" and thus could not be talked about. Now, of course, it's accepted as valid.
Let me discuss reviewer bias. When a bunch of experts are invited to discuss a particular topic, who gets invited to the panel? The ones the empaneling authority agrees with? How can that be unbiased? So, even though a large number of scientists were around a table, that does not mean that they were unbiased. It also does not mean they were biased; we simply do not know.
Big tobacco was an expert at creating concensus, and we we know exactly how biased their panels were. We also know their motive: money.
However - what reason could the National Cancer Institute (NCI) have for promoting such a bias, if one were to exist? Is it likely that the NCI would empanel a biased set of reviewers? What would be their motive?
I don't like conspiracy theories. And I'm not going to advance one here. I hope that someone would blow the whistle on that kind of thing.
I haven't looked at the studies. I haven't looked at the questions the researchers were asking. So I don't really know whether or not to trust them. But that many studies came to the same statistical conclusion indicates that I should, at least partly. (Many reviewers, many viewpoints, same conclusion... okay; we'll go with it for now)
However, statistical studies control for other causes; that is, if one activity that's known to be a risk factor is factored out, other things that cause that activity are hidden from view. Therefore, in order to know what's going on, we need to look at a different question.
This site raises questions about these studies, by providing biological evidence for the site's position. This site is clearly biased. Does that mean they are wrong? I'm not so sure. You see, we know that childlessness increases the risk of breast cancer. It is not the abortion that's the risk factor, but childlessness that increases the risk of cancer (one of four major factors). When abortion is the cause of childlessness, well, you'd think that'd be a conclusion. But, actually, that's not the case; people who've had children also have had abortions.
Here is the press release about a study that links abortion and breast cancer.
Hm. Okay. We have conflicting studies and conflicting evidence. Who do we believe? The biggest? The mostest? Maybe not. We can ask a different question, though.
The ACS reports a link between reduced risk of breast cancer and childbearing at an early age.
After looking at all of this mess, I must conclude that many scientists and doctors are not looking at the right things or asking the right questions. My question would not be, "Does abortion cause breast cancer?" but rather, "What are the risk factors associated with breast cancer?" Just as I should ask, "What are the risk factors associated with lung cancer?"
Causation is hard to identify; risk factors are easier.
Abortion probably does not cause breast cancer (in the face of millions of cases reviewed, I will not claim it does). However, people should be aware of what the risk factors are. There are four primary risk factors associated with breast cancer:
1) Childlessness;
2) Small family size;
3) Little or no breastfeeding; and
4) Having a late first full term pregnancy.
What to conclude about abortion? That it was not statistically linked to breast cancer means that abortion is evenly distributed among the population of women of childbearing age.
To use another example: spending money on cigarettes is connected to lung cancer, but does not cause it. It's not the spending that's the problem; it's the smoking. In a similar way, abortion is connected to breast cancer, but does not cause it.
To announce that abortion is not linked to breast cancer is to tell a truth - but not the vital truth concerning the risk factors of breast cancer. In a sense, to vaunt the nonconnection is to deceive the listener or reader. If small family size is a risk factor for breast cancer, and if choosing abortion is the cause of small family size - you can see the connection, even though abortion is not a risk factor itself. That does not mean that choosing not to have a family is wrong. It just means there are risks associated with that decision. But to have a larger family can also mean higher risk. Balancing risks is part of being human. There are risks associated with all parts of life, even in seemingly non-risky behaviors.
The choice of whether to abort or not is a decision best reserved for the woman who is pregnant. That woman should be well-informed of the risks, benefits, and results of that choice, whichever way she chooses, and she should not be pressured one way or the other by the health-care professional who is offering the choice (that would be unethical).
I hope that made sense.