Babs Kay is on her ABC (abortion = breast cancer) bullshit soap box Again. (That's a Do Not Link link, which I am going to use for all sites I don't want to get hits.)
This time she's citing a meta-study out of China -- where, you know, the health care system, abortion regime, environment, etc. etc. is exactly like North America.
Problem with this study is, as usual, its methodology, not to mention the obvious flaws in trying to compare these societies.
Short version: same old crap dressed up yet again as The Game Changer !!!!!!!!
Um. No.
Let's hear from an epidemiologist with the American Cancer named Susan Gapstur.
"The findings of this meta-analysis should be viewed with caution," Dr. Susan Gapstur told The Wire in an email. Gapstur is the vice president of epidemiology at the American Cancer Society. She notes that almost all of the studies cited in Dr. Huang's analysis used something called the case-control method, which tends to produce misleading results. In the case of the abortion-breast cancer link, women with breast cancer who self-report their reproductive histories tend to do so more accurately than women who are cancer-free. And in countries like China, where abortion still carries a significant stigma, that "recall bias" can be reinforced. "This 'recall bias' can make it look like breast cancer is associated with abortion when it is not," Gapstur explains. Case-control methods, it should be noted, have produced links between breast cancer and induced abortion before.Further, the most reliable studies in the mega-study do NOT find a link.
All but two of the studies included in Huang's analysis used the case-control method. The remaining two were prospective cohort studies, which track women over time, instead of relying on self-reported historical results. Those two studies, Gapstur notes, did not find a link between abortion and breast cancer. In fact, the eight studies that appear to be the most reliable of the group found no link between induced abortion and breast cancer risk. "The association only became apparent as the quality of the studies decreased," Gapstur told The Wire, noting that some of the included studies were not published in peer-reviewed publications. In other words, the work might not be vetted by independent professionals in the field.
Over at Lifeshite, Joel Brind, granddaddy of ABC himself, calls Dr Gapstur an 'epidemiologist'. Gee, I dunno, she looks pretty well qualified to me.
I'm sure you are as bored by this as I am, but I take it as a duty to call them out. Every. Single. Fucking. Time.
And, really, Babs should be ashamed of herself, peddling this fear and worry to women year after year, despite getting regularly smacked down for it.
ADDED: From Babs herself in the comments to her piece:
Barbara Kay
Barbara Kay is an opinion journalist and it is her job to present the side she finds more persuasive. Other opinion journalists are free to adduce evidence they feel is more persuasive. Let the debate go on.
Or, shorter: Babs Kay is free to make any shit up she wants without regard for normal journalistic standards of accuracy and truth.
ANOTHER ADDITION: A media release on the anti-choice book, Complications, that Babs touts in her column. And oh look who's giving it a rave review. Our old pal, serial research-fudger, and perfesser of home ec, Priscilla Coleman. Surprise, surprise.
5 comments:
I've been looking at this study and plan to write my own analysis of it. I have a PDF of the whole study if you want it.
Basically, the meta-analysis is completely worthless because of the "recall bias" effect in most of the studies included. The authors state at one point that there's little abortion stigma in China, then contradict themselves later and say there's still a lot of stigma, especially for 2nd and 3rd abortions.
Although that one factor is fatal to the meta-analysis, there are several other major problems with the study - each of which is also fatal to their agenda of trying to show a causal link between abortion and breast cancer. Actually, it's a wonderful example of B.A.D. science falling on its own sword in spectacular fashion. The whole study is shoddily presented and badly written. "Prestigious journal" my ass!
One further example is that the authors never mention any other possible explanation for a rise in breast cancer in China over the past 2 decades - but there are several contenders far more likely than abortion - increasing affluence with changes in diet and increased body weight (latter is linked to an increase in breast cancer), increasing industrialization and exposure to environmental toxins, and increased testing for breast cancer with resulting increases in diagnoses.
Also, there is no correlation between breast cancer and abortion in many other countries, such as western Europe, Russia, etc. You can't just pick one country like China and focus only on it. Out of 195 or so countries in the world, some will inevitably show an association between abortion and breast cancer, just by chance, while many other countries won't. There's so many factors at play, that even if abortion was a small causal factor, it would be drowned out by others. But of course, an association does not mean causation.
Great work, Joyce. As you said, a splendid example of B.A.D. science.
I wonder if there was any "enticement" for the Chinese researchers to produce this. Some deep pockets on the anti side. I sure wouldn't put it past them.
Me neither. It's hard to say without more info, but given that Joel Brind has been promoting it and publishing summaries of it, I would guess he's been in cahoots with the study authors for some time. They also over-rely on his 1996 meta-analysis, citing it several times like it was gospel, even though experts dismissed it at the time as GIGO (garbage in, garbage out).
Anonymous submitted the same comment twice (all ready to go on on a clipboard?) with a bunch of assertions. Then tried to back it up with a link to the Breast Cancer Prevention Institute.
Ha. No dice. That is a bogus, BAD science outfit, founded by Joel Brind himself.
Go peddle your lies elsewhere, Anon.
Post a Comment