Talk:Abortion–breast cancer hypothesis/Archive 2

Archive 1

Archive 2

Archive 3

Archive 4

Archive 5

opinions of a pro lifer

given my obvious bias my opinions can hardly be taken as POV r watever but this article does seem to concentrate much on disproving rather than a clear balanced approac i cite many soureces as to the connection;

(lengthy cut-and-paste from http://www.abortionbreastcancer.com/abc.html removed)

next the actual cause is explained;

(another lengthy cut-and-paste from http://www.abortionbreastcancer.com/abc.html removed)

i could go on.....i dont expect all of this to be included.....but i do expect in the interest of POV that prolifers be respected and akknowledged for their genuine scientific research....and that this "tactic" not be labelled as scaremongering but as a serious effort to expose the actual risks involved in abortion....understandbly this encyclepdeia has to remain balanced but from my point of view (POV) this article is highly biased (unlike many of the other articles on abortion) again i reiterate that akknowledgement be given to the overwhelming research into this connection and much of the bias and negative references to prolifers be removed in the interest of NPOV which seems to be very obvious in this article.....--83.49.150.155 17:21, 7 February 2007 (UTC)

The article reflects the fact that every major medical and scientific group that has studied the question has found no link between abortion and breast cancer. Many on the pro-choice side have referred to the pushing of a supposed link, in the absence of such evidence, as "scare-mongering". These things are noted in the interest of presenting all sides with appropriate weight. If anything, this article gives far too much weight to the views of a small fringe of the scientific community whose views are widely felt to be unsupported. MastCell 17:53, 7 February 2007 (UTC)

• the quotes I see have been removed....do they show evidence of a link or not...? if not where is the scientific error (i was sure i presented the biological reasons that abortion caused breast cancer again where is the error), i would be personally intersted as the quotes seemed to show large consensus (rather than a "small fringe"), and if so why were they removed? (granted they were long and tedious) it obviously hasnt been found by "every major medical and scientific group" when the quotes i gave were at least some of them from major scientific groups....
• if the page is going to present the opinion of a "minority" (which i also challenge) at all then it should at least mention why and how this logical conclusion was made, (ie the biological evidence and scientific research which has yet to be disputed) rather than dismissing their views as "scaremongering".....

--83.49.150.155 19:46, 7 February 2007 (UTC)

This isn't really a forum for discussing the science or re-creating the dispute - please see the talk page guidelines. ABC proponents are a minority. No major medical organization has concluded that an ABC link exists, and the NCI specifically rejected such a link. The reasons why this conclusion was reached, as well as references to most of the appropriate primary and secondary sources, both for and against such a link, are cited in detail in the article. The neutral point of view policy and associated FAQ would probably be worth reviewing. MastCell 20:34, 7 February 2007 (UTC)

if this article isnt "for discussing the science or re-creating the dispute", and if i accept all of the above, it still doesnt justify the negative connations towards the pro life movement. The hypothesis is a valid theory (if not proven to be so, although the wesite i used wasvery convincing to me, using doctors and citing various studies and showing graphs of the increase), therefore it is clear that the pro lifers are not trying to "scaremonger" women into haveing an abortion but are trying to suggest that there may be certain negative affects to abortion that may harm the mother....this is out of genuine concern for the woman and should not be viewed as mallicious which is what this article suggests...

and I just noticed the link to the website I referenced, so apolagies, as i thaught i was brining something knew to the article, obviously it has already been reviewed —The preceding unsigned comment was added by --83.49.150.155 21:25, 7 February 2007 (UTC)83.49.150.155 (talk) 21:13, 7 February 2007 (UTC).

Indeed the webpage in question has been looked at; but for such a controversial topic there is conscious effort to source as much as possible from scientific studies directly. This helps bypass bias that might exist in secondary sources, such as abortionbreastcancer.com. As to pro-life intentions both sides of that issue are presented in the article; but it can be verifiably said some zealous pro-lifers are not beyond scare tactics to influence women. - RoyBoy ⁸⁰⁰ 22:07, 7 February 2007 (UTC)

• to clarify....u are correct in saying that its best to cite primary rather than secondary sources, however abortionbreastcancer.com's references primary scientific studies, and would it not be prudent to investigate where these references come from and then you may find that they are verifiable, and hopefully u might be able to cite them as primary sources (obviously many of the sources are taken from books or reports by reputable doctors but that shouldnt b a problem either).....

it also can be verifiably said some zealous pro-choicers are not beyond scare tactics to influence women (and yet this isnt hinted at, at all)....i cite (unfortuanatly havnt time to go find the reference again, although im certain i read it on the internet) a case in the States where a women was bullied and threatened (by the local feminist organisation) because she decided to go through with her pregnancy after being raped.....there are many other cases....and there are many women who are pressured into having an abortion rather than it truly being their "choice"

this is beside the point....either the article is balanced and cites example of "zealousness" on both sides and their tactics....or it is a nonPOV article which deals with the merits or demerits of this particular hypothesis (and not with the overall debate on abortion which is dealt with elsewhere) —The preceding unsigned comment was added by 83.45.178.242 (talk • contribs).

Those aren't equivalent examples since they are anecdotal (restricted to individuals). Furthermore as you've pointed out, it has nothing to do with the ABC issue. Use of the ABC issue as a scare tactic is relevant to extreme pro-life elements; extreme pro-choice elements call the ABC issue "pseudoscience". They too are mentioned in this article (in the lawsuit section); so both extremes of the ABC issue are represented in the article. Now, it is possible with very careful referencing to provide a section called "Pro-choice bias"... but its something I'd have to collaborate on with other editors. - RoyBoy ⁸⁰⁰ 23:53, 8 February 2007 (UTC)

Lead tweak, pro-choice section and good article

I've tweaked the lead to clarify ABC wasn't invented by pro-life groups. However, I am trying to figure out if "in humans" is appropriate prior to pasting lead changes to Abortion. Looking at the Russo and Russo research and trying assess if they actually concluded anything. Found out I only have 2 middle pages of the 1987 study, I need to get a full copy of it to see their conclusions. The thing is, Russo and Russo conducted three studies... each more detailed than the next; so it seems to me they verified their rat findings and hypothesis; in rats. So I think "in humans" is accurate, but might be too leading.

So as to the pro-choice section, I want this section included as a way to help the article achieve my next topic...

...good article status. Do my lead tweaks, and proposed pro-choice bias section improve the article and take it to where it needs to go for good article status?

Just for future reference the Pro-choice bias section would compile existing information and sources:

• pseudoscience quote from a pro-choice lawyer (juxtaposed with conflicting scientific data)
• poor pro-choice research:
  • Women's eNews article (in Joel Brind)
  • Joyce Arthur's article (in Joel Brind)
  • weak research 1989 cohort used by Planned Parenthood
• maybe move Daling's quote there

So, yeah... thoughts? - RoyBoy ⁸⁰⁰ 04:16, 22 February 2007 (UTC)

As you know from Talk:Cancer, my opinion is that it's always important to draw a clear distinction between rat/animal studies and human data. I would definitely keep the clear statement that the link has not been verified in humans, and is not recognized by any major medical organization, in the lead. If anything, mentioning Brind in the lead is probably undue weight, given that he was 1 out of >100 experts and had a very clear investment, as a major proponent of the ABC link. But that's not too big of a deal. I guess a "pro-choice bias" section would be OK, if there's really evidence of a bias, but perhaps it would make more sense to change the "Pro-life bias" section heading to something a little less inflammatory. Really, when the medical data, NCI consensus, etc are all pretty clear that there's no link, then "bias" comes into play primarily when examining the motives of those who disregard the scientific consensus and keep pushing a link. But again, that's my 2 cents. It's a good article. I'd be interested in outside opinions as to whether it's a good article. MastCell 06:15, 22 February 2007 (UTC)

My bad in communication regarding "in humans". I'm concerned that including that statement actually makes the ABC association stronger than is desirable; because if you say "not verified in humans", then that leaves the next question... so it is confirmed in other species!? Even inferring strong evidence in other species; even though it is accurate... makes me uneasy. If you take "in humans" out then "not verified" can be a broader statement that could, if the reader chose, include not verified in rats too.

As to Brind, absolutely undue weight in the context of the NCI workshop, however it does provide a way to introduce the leading proponent and at the same time provide insightful context on his position vs. scientific consensus. Hence a lead mention seems appropriate, perhaps even necessary... though you are getting me to reconsider; after all it can be moved to the NCI workshop section which is now right below the lead (thanks to Sophia's suggestion of an article inversion) and 1 vs 100 experts can remain without naming him immediately. Hmmmm... have to sleep on this, really want to get the lead optimized.

While bias is certainly primarily on the pro-life side; it is key to cover the dismissive bias of others, which I've already done with the sources already included in the ABC/Brind articles. I could hit two birds with one stone by tweaking "Pro-life bias" section to "Bias" and include both in one section. Think that would be ideal? - RoyBoy ⁸⁰⁰ 04:23, 23 February 2007 (UTC)

Thinking more about Brind in lead before I post... I think I'd base my opinion on whether or not it is appropriate for Brind to be mentioned in Abortion given the ABC lead is replicated there. Trying to formulate a matrix of pros and cons... my instinct is Brind isn't notable enough to be in the Abortion article. LOL, just realized this is analogous to the DCA mention in Cancer. I'll change it now, since otherwise it will keep me awake... and I really need my sleep these days. But I won't replicate in Abortion until you give me a nickle. :"D - RoyBoy ⁸⁰⁰ 04:23, 23 February 2007 (UTC)

That took longer than I thought, tried to remove redundancies, improve flow and end it with your balancing sentence. - RoyBoy ⁸⁰⁰ 04:53, 23 February 2007 (UTC)

Haven't had a chance to read it in depth yet, but your edits look good. I like the idea of changing "Pro-life bias" to simply "Bias" (or "Politicization" or something?) and dealing with both pro-life and pro-choice political appropriations of the issue. Re the rats, you could say that a handful of studies have suggested an association in rats, but studies have not found a similar association in humans... er something. MastCell 20:01, 23 February 2007 (UTC)

Yeah, I was thinking Politicization would be a good option too; I'll do that soon. I guess I'll leave the "in humans" as it is and replicate to Abortion. Let the inferences go where they may. - RoyBoy ⁸⁰⁰ 21:11, 23 February 2007 (UTC)

Discover article

True Discover isn't officially pro-choice, as a progressive science magazine it isn't a stretch to think it is pro-choice. Moreover, when I decided to put it in the pro-choice section, it was because of its author, its content and this other article.

Regarding its content, I find this paragraph particularly troublesome and lacking scientific rigor, or more specifically, lacking fact checking:

Brind dismisses response bias as an unproven hypothesis, but others have found ample evidence for it. In Sweden, epidemiologist Britt-Marie Lindefors-Harris of the Karolinska Institute took advantage of her country's nationwide registry of legal abortions. In a project documented in the American Journal of Epidemiology, Lindefors-Harris conducted a case-control study of abortion and breast cancer, but with a twist: She checked government records to see if the participants were telling the truth about their reproductive histories. Many of them, it turns out, were not. Out of 829 women, 29 appeared to misrepresent their abortion history, with the vast majority of underreporting coming from healthy women in the control group. Based on those numbers, Lindefors-Harris calculated that "an observed increase in risk of up to 50 percent may be caused by response bias."

I strongly feel it should either be put back into the pro-choice section, or removed from the external links section entirely, as it can remain as a footnote reference. Another reason I even linked to the article was to balance the sub-sections out to 5 links each. - RoyBoy ⁸⁰⁰ 16:51, 11 March 2007 (UTC)

I see the issue with balancing the number of links. The author does appear to be pro-choice and to write from that perspective. However, there is a difference between Discover (a scientifically focused magazine) and Mother Jones (which has a clear political perspective). The rest of the "pro-life"/"pro-choice" links appeared to be from partisan groups, so Discover did not seem to fit; it seemed to fit best under "Scientific". But I don't feel especially strongly about it, and it sounds like you do, so if you'd like to put it back I won't object. MastCell 20:34, 11 March 2007 (UTC)

I agree with your perspective, that's why I think removing the link completely might be the best option. Although I'm confident in my initial categorization of the article, it could successfully be argued it isn't our place to make such nuanced determinations. Even if they are correct; I'll put it back to pro-choice, but something tells me this won't be the last time I address this subtle issue. I'm going to remove the new pro-life Canadian link as non-notable, and balance the links out again. - RoyBoy ⁸⁰⁰ 23:18, 11 March 2007 (UTC)

Citation format

I was thinking of converting all the PubMed references to use the {{cite}} template format. Any thoughts/objections? MastCell 20:39, 11 March 2007 (UTC)

Go for it, it was simply referencing exhaustion on my part that older links weren't converted. I did the vast majority of the refing after a peer review prior to finding out about automated templates; I wasn't pleased to say the least. - RoyBoy ⁸⁰⁰ 23:19, 11 March 2007 (UTC)

It won't take long; there's an automated tool that helps you do it (see my user page, at the bottom). I'll take care of it in the next few days. MastCell 00:11, 12 March 2007 (UTC)

I've started the process - I'm pretty careful, but you may want to double-check the new links to make sure they look right. MastCell ^Talk 16:30, 18 March 2007 (UTC)

Partiality

I think this article violates the premise of being neutral - it focus too much in the abortion debate, and neglects the cancer debate. Why it's not part of a cancer debate? What are the other causes (proved, disproved, and under dispute) of cancer? Do cigarettes cause cancer? Do cell phones cause cancer? Do bras cause cancer? Does the Internet cause cancer? Albmont 17:41, 27 March 2007 (UTC)

The topic is one that has become part of the abortion debate, at least in the U.S. If you're looking for information about other causes of cancer, the main article on cancer is a good starting point. MastCell ^Talk 18:43, 27 March 2007 (UTC)

This topic is mentioned in breast cancer as unproven. I think what Albmont is getting at; is the lack of Cancer categorization. Albmont has a point, though I don't know where this would fit in the Cancer category structure. - RoyBoy ⁸⁰⁰ 23:21, 28 March 2007 (UTC)

My advice would probably be to swing for the broadest categorization possible and place it under something like Category:Oncology. -Severa (!!!) 23:50, 28 March 2007 (UTC)

That would be the category of choice currently, its either that or create a sub-cat as discussed below. - RoyBoy ⁸⁰⁰ 03:42, 29 March 2007 (UTC)

And that maybe there should be a Cancer article covering the various claims/causes of cancer that are currently in dispute. For cell phones for example, its under Mobile phone radiation and health, but it could be helpful to centralize borderline causes of cancer into one article? - RoyBoy ⁸⁰⁰ 23:39, 28 March 2007 (UTC)

Maybe something like "Disputed risk factors for cancer"? Don't know how many articles would be in the category though - I mean, is secondhand smoke still a "disputed" risk factor just because a few friends of the tobacco industry are still arguing? Is asbestos "disputed", or radon? Might be opening a can of worms. MastCell ^Talk 00:04, 29 March 2007 (UTC)

I tend to agree that the creation of a "Disputed risk factors for cancer" category could become problematic. You could always winnow out POINT-ish categorizations on a case-by-case basis but this could just be making more work than is actually necessary. And per WP:CAT, "Categories appear without annotations, so be careful of NPOV when creating or filling categories. Unless it is self-evident and uncontroversial that something belongs in a category, it should not be put into a category.". -Severa (!!!) 03:03, 29 March 2007 (UTC)

A little confusion there, I was contemplating a disputed cancer article; though a category would be much easier to manage. That's a great idea Severa! :"D No need for replicating content in that scenario. - RoyBoy ⁸⁰⁰ 03:26, 29 March 2007 (UTC)

I mean, good idea MastCell. I need more sleep and less Wiki; or just more of both. ;'D RoyBoy ⁸⁰⁰ 03:30, 29 March 2007 (UTC)

Problematic would be an understatement... but then again I'm use to problematic topics. Regarding second hand smoke and lung cancer, I consider myself slightly reserved on that issue, and found a great article by a straightgoods kind of column (tried to find it with no success) which examined the rushed EPA study with 90% confidence interval; and commented that for once tobacco lobbyists might be right... though those were early days in the opening of research. At least the initial evidence was rushed a bit to follow political/social trends.

I'll have to discuss the cancer article issue with the relevant wikiprojects and try to assess the mood for such an article, and if it would be an asset or detriment to Wikipedia. - RoyBoy ⁸⁰⁰ 03:23, 29 March 2007 (UTC)

Update, creation of a disputed cancer category, not an article. - RoyBoy ⁸⁰⁰ 03:29, 29 March 2007 (UTC)

Here's an speech by Crichton going into a bit of detail on the second hand smoke scare. - RoyBoy ⁸⁰⁰ 03:37, 29 March 2007 (UTC)

Crichton went off the rails a long time ago is a professional contrarian these days, IMHO. The interesting thing about secondhand smoke is that, thanks to the tobacco settlement, you can peruse the once-secret memoranda of the tobacco companies, in which they plotted strategy to manufacture uncertainty about the EPA report and go after the EPA's credibility. "Doubt is our product," as one of the memos said - a beautiful turn of phrase. Regardless of that original EPA report (which I haven't looked into in too much detail), their findings have since been confirmed by (what I consider to be) an indisputable amount of accumulated evidence. But I digress. Sure, why not see what others have to say. We should be careful to make the category title as general and uncontroversial as possible - categories have a nasty habit of becoming a vehicle for tendentiousness. MastCell ^Talk 03:42, 29 March 2007 (UTC)

Indeed, hardly surprising given the stakes involved. I haven't looked closely into the subject in years, but uncertainty is easy to foster when confidence intervals are tweaked. - RoyBoy ⁸⁰⁰ 22:20, 31 March 2007 (UTC)

USAToday article link

Here is an article from USAToday about the abortion-breast cancer link if you want to include it in the article.[1] --Nehrams2020 23:59, 23 April 2007 (UTC)

Thanks for that - we should work in both the original study (from Archives of Internal Medicine) and perhaps this article or some other popular-press coverage. MastCell ^Talk 01:59, 24 April 2007 (UTC)

Nice recap of the current consensus, but nothing new; and the level of reporting is slipshod on report bias, which carries forward an assumption from the researcher that response bias is statistical significant... par for the course on the ABC issue. - RoyBoy ⁸⁰⁰ 20:13, 24 April 2007 (UTC)

Made use of it in the conclusion section. - RoyBoy ⁸⁰⁰ 20:20, 24 April 2007 (UTC)

I think we should cite the study itself, as well as PMID 16646050 and PMID 17111259, as recent studies which have further supported the consensus that there is no link between abortion and breast cancer (I can't find them cited in the article thus far). There does seem to be an accumulation of reproducible evidence from a number of different groups supporting the absence of a link, and this isn't necessarily reflected in the article. I guess mentioning the protective effect of abortion against uterine cancer is probably too much, though... MastCell ^Talk 21:45, 24 April 2007 (UTC)

I'm concerned with the word "reproducible". Uninspired science is as reproducible as groundbreaking science. My current thought is merely regurgitating the current consensus (unsupported insofar as response bias is concerned) just isn't notable. Reference it, sure... mention it, I can't see why. There are significant problems with previous large record based ABC studies finding no link; while I have not looked at these newer studies, nor do I intend to, I cannot help but think these "unbiased" record based studies are quite capable of having problems of their own.

As to reduced uterine cancer, I think that should definitely be mentioned, if this is the only study to mention it, then with the caveat unconfirmed or tentative... perhaps put it in a confounding factors section for now, and it could theoretically be the beginning of a new section "Benefits" if other positive anti-cancer effects to abortion could be turned up. - RoyBoy ⁸⁰⁰ 00:22, 25 April 2007 (UTC)

Sure, but reproduceability of results is one of the cornerstones of the scientific method. Every individual study has problems - but when individual studies by different groups, in different populations, reproduce similar findings, then we can be more and more confident that their findings are correct. As to uterine cancer, I have not seen that mentioned in any other study, although I haven't read through all of them in their entirety. Perhaps mentioning it as "One study found that..." would be appropriate. MastCell ^Talk 01:05, 25 April 2007 (UTC)

Indeed, but I'll reiterate more for lurkers than for you; that broader sociological factors come into play for scientific paradigms/history. Science unfortunately isn't immune to group think, wishful thinking and political correctness. It is these realities, which I strongly feel I've confirmed in the ABC issue, and exist in previous ambiguous/debatable issues (ie: The Blank Slate) where the consensus encouraged — and was distinctly uncritical — of unchallenging science which reinforced the consensus in a negative feedback loop.

My trust/faith in science wants/needs it to be better than that. I refuse to concede anything unless the ABC biological mechanism is taken apart and examined. It needs to be explained how a significant hormonal event (early pregnancy) cannot influence breast cancer; when we know for a fact hormones specifically impact breast cancer. As long as science continues to beat around the bush; epidemiology can stick its head in a bucket for all I care. ;'D Yeah, I'm in a sour mood... need to catch up on some sleep. Night MastCell... now that I think about it, MastCell is a sweet handle. - RoyBoy ⁸⁰⁰ 02:36, 25 April 2007 (UTC)

Thanks... hopefully that's not an indication that I'm causing you metaphorical irritation and anaphylaxis. I agree that science is ultimately a human endeavor, meaning it's not immune to broader social/political influences or groupthink. Or, for that matter, contrarianism (a la Brind and Michael Crichton). Your points are valid and well-taken; however, I think our task on Wikipedia is simpler than resolving these thorny issues. We just need to present scientific opinion with an appropriate amount of weight based on its current acceptance in the scientific community. That's my 2 cents, anyway. Hope you feel better after some rest and that I'm not irritating you unduly. MastCell ^Talk 05:11, 25 April 2007 (UTC)

Nah, you're coo. I have more of an issue with semi-snorers.

As to your addition, its certainly short enough, but screws up the narrative. "Critics of these studies argue they are subject to selection bias" isn't necessarily applicable to the cohort studies you've listed; also you've created a repetition of the word "prospective". I'll merge your addition into the prospective sentence in the third paragraph; and wait on your feedback before replicating to abortion.

We agree on Wikipedia's role. - RoyBoy ⁸⁰⁰ 23:48, 25 April 2007 (UTC)

I think it looks good. MastCell ^Talk 04:15, 26 April 2007 (UTC)

Further cohort studies

Good section, tweaked out repetition. I'd also like to see the number of case patients found and emphasized; all these overall large numbers mean squat to me; and can be very misleading to lay readers. Removed following, as I like to avoid the impression these studies dealt with the subjects in a direct way or somehow were carefully ongoing for many many years; rather than the more likely scenario of data crunching an available health database. - RoyBoy ⁸⁰⁰ 23:57, 26 April 2007 (UTC)

Is this article stable enough for a GA review?

A diff over five days seems to indicate large content changes, (mostly additions it seems) [2] and I had intended to review this article, until a bump dislodged my laptop hard drive, and I figured I might as well check the history anyway after it restarted. Homestarmy 19:50, 30 April 2007 (UTC)

I've started my review now, I really want to go all out on this one, i'm only partly done, but it may take awhile. Homestarmy 18:54, 1 May 2007 (UTC)

Made a punctuation change, and a tone change to one sentence. Hopefully it does not put you out. - RoyBoy ⁸⁰⁰ 22:55, 1 May 2007 (UTC)

Haven't gotten down to the conclusion yet actually, this thing is a big article, and more complicated than I thought it would be.... Homestarmy 23:40, 1 May 2007 (UTC)

ABC GA leftovers

• Politicization: When pro-life advocates link abortion to breast cancer, some claim that the goal is to stop women from having an induced abortion. Because breast cancer elicits disproportionate fear in women, there exists the concern that pro-lifers use it as a scare tactic. Abortion-breast cancer advocates who oppose abortion have been accused of focusing on positive and/or averaged results, ignoring caveats and low-risk subgroups. These advocates rebut by stating that their ABC information is for the benefit of women's health and to provide informed consent.

• NCI workshop: It should be noted that Dr. Brind himself was invited and that federal agencies are the major source of research funding in the United States, making it inherently difficult to avoid inviting government funded scientists.

• Confounding: It should be noted the overall incidence does not effect ABC studies with proper controls because the case and control subjects would be equally affected.

• State laws: These state laws put up further barriers to elective abortion, and

• Conclusion: No major international cancer organization considers abortion to be a breast cancer risk factor despite the statement by pro-life advocates and some medical researchers that there is a link. While some retrospective studies and animal data have found evidence for a link, some larger international prospective cohort studies have found no evidence that a link actually exists. The scientific consensus garnered from recent large studies is that positive results seen in earlier studies occur because of [response bias], and that if this alleged bias were removed there would be no significant abortion-breast cancer association.
  Despite a scientific consensus that there is no link, conflicting studies and even contradictory results within studies exist,. This contradictions may have occured because the effect being sought is too small, does not exist, and/or as a result of incomplete data and flaws in the studies. Most abortion-breast cancer correlations are minor when compared to established genetic and lifestyle risk factors for breast cancer. It is also worth considering that even if the data were interpreted to suggest that a statistical correlation between abortion and breast cancer does exist, correlation does not imply causation. As with other politically charged topics, the science and the uncertainties inherent in the scientific method are in danger of manipulation and exploitation for political ends by either side.

---

Above are things that were removed from the ABC article because of the GA review. If they can be better referenced and/or reintegrated into the article; that would be great. - RoyBoy ⁸⁰⁰ 01:29, 4 May 2007 (UTC)

A note to self, removed: Brind also asserts that many invited scientists had a conflict of interest because they were dependent on the NCI or other federal agencies for grants.

Might be re-added to political section. - RoyBoy ⁸⁰⁰ 04:46, 21 November 2007 (UTC)

ABC as political tactic

The following as further evidence that ABC is a political tactic, not a medical or scientific "debate":

• http://www.washingtonmonthly.com/features/2004/0410.mooney.html

It is therefore appropriate that the first sentence of this article makes it clear that this is a concept perpetuated by pro-life activists.

I also see no problem with the following paragraph. It is impeccably referenced.:

No major international cancer organization considers abortion to be a breast cancer risk factor despite the statement by pro-life advocates and some medical researchers that there is a link. While some retrospective studies and animal data have found evidence for a link, some larger international prospective cohort studies have found no evidence that a link actually exists. The scientific consensus garnered from recent large studies is that positive results seen in earlier studies occur because of [response bias], and that if this alleged bias were removed there would be no significant abortion-breast cancer association.
Despite a scientific consensus that there is no link, conflicting studies and even contradictory results within studies exist,. This contradictions may have occured because the effect being sought is too small, does not exist, and/or as a result of incomplete data and flaws in the studies. Most abortion-breast cancer correlations are minor when compared to established genetic and lifestyle risk factors for breast cancer. It is also worth considering that even if the data were interpreted to suggest that a statistical correlation between abortion and breast cancer does exist, correlation does not imply causation. As with other politically charged topics, the science and the uncertainties inherent in the scientific method are in danger of manipulation and exploitation for political ends by either side.

--IronAngelAlice 21:27, 1 October 2007 (UTC)

Not necessarily; and the first sentence does clarify it; however, they are not the only ones "perpetuating" ABC, some pro-choice scientists support the issue. Please refrain from quick edits based on one article you came across. While I agree with your characterization of the ABC issue; I disagree with the prominence you ascribe to it. ABC being a pro-life topic clearly stated in the Lead of the article. - RoyBoy ⁸⁰⁰ 21:57, 1 October 2007 (UTC)

What you are suggesting is that the Washington Monthly article is wrong. Do you have evidence that "Pro-choice," or preferably "neutral" scientists support the "issue."? If not, I suggest it should be made clear that ABC is a hypothesis being used by pro-life advocates as a political tactic, and has little to do with public or women's health. --IronAngelAlice 22:02, 1 October 2007 (UTC)

It is wrong if it states: "is a concept postulated by pro-life activists." In actuality it was hypothesized by Russo and Russo; and pro-lifers have picked up on the limited evidence and ran with it. That doesn't mean pro-life created it. As to evidence for neutral scientists, you could actually read the article prior to editing it. Howe and Daling come to mind.

Your reference is interesting politically, but isn't that good on the science. For example "biggest coup" implies that changing to NCI website to say the science is inconclusive makes it appear saying that is inaccurate. It frames it in such a way as to state an opinion as fact. The opinion, based on the NCI workshops conclusions isn't rare is maintream... but the workshops shortcomings are noted in this article. Also of note, the studies they included/excluded in their review were not made public... making the politically convened workshop not open to review or scrutiny. It was convened in response to political changes, it doesn't take much to imagine it may have been organized by those opposed to Bush, pro-life etc. I don't mind that much persay, but it doesn't exactly create a neutral atmosphere. - RoyBoy ⁸⁰⁰ 22:40, 1 October 2007 (UTC)

I added your ref to the lead. - RoyBoy ⁸⁰⁰ 22:44, 1 October 2007 (UTC)

It is certainly wrong to indicate that the hypothesis was initially proposed by pro-life activists, or that every last supporter of it is pro-life. Nonetheless, we should perhaps more explicitly acknowledge (as the Mooney ref actually says) that at present, the hypothesis owes whatever prominence it enjoys to a concerted political effort by pro-life activists rather than any real scientific currency. MastCell ^Talk 23:43, 1 October 2007 (UTC)

It already says that in the third paragraph:

"The ongoing prominence of the abortion-breast cancer hypothesis, despite the lack of supporting scientific evidence, is seen by some as a part of the current pro-life "women-centered" strategy against abortion."

I'm very concerned "despite the lack of supporting scientific evidence" is inaccurate, as it seems to imply ABC is scientifically unsound. It isn't, rather it is unsupported. - RoyBoy ⁸⁰⁰ 00:18, 2 October 2007 (UTC)

To clarify, "lack" is the problematic word. The ABC issue is not lacking science indicating a correlation. Even the largest study, Melbye, contains evidence to that effect within it. - RoyBoy ⁸⁰⁰ 00:58, 2 October 2007 (UTC)

OK, but now we're getting back into our interpretations of the data and verging on WP:OR. So far as I know, the major scientific groups that have looked at the question have found that the ABC link, in humans at least, is unsupported by scientific evidence. Perhaps there are scattered findings that suggest a link, but when the scientific evidence is viewed as a whole by reliable sources and experts in the field, they've felt that the hypothesis is not supported by the evidence. If you'd like to remove the word "lack" and rephrase as "unsupported by the scientific evidence", then I don't have a problem with that. MastCell ^Talk 02:57, 2 October 2007 (UTC)

That's unfair, it isn't verging on OR to note in the Melbye study that breast cancer risk went up 3% for every week of gestation. This isn't "scattered findings", it only appears as such because of how the ABC issue is framed by pro-choice sources. The preponderance of interview studies indicating a link isn't scattered either, and the allegations of response bias are just that... allegations. No one has backed up the criticism with actual science to show its a statistically significant factor. It's just taken for granted by researchers, reporters etc. that these studies have been discredited. That simply is not the case. - RoyBoy ⁸⁰⁰ 21:15, 2 October 2007 (UTC)

No, it's how the ABC issue has been framed by (among others) the National Cancer Institute and the American Cancer Society. Those are not "pro-choice" sources. No one's talking about "discredited", which is a strong and inflammatory word; the evidence in favor of a link is simply strongly outweighed by the evidence against one. MastCell ^Talk 21:27, 2 October 2007 (UTC)

I inferred discredited from scattered, my bad and I apologize.

The NCI and ACS among other reliable sources are exactly the ones choosing to minimize positive results (with sound and unsound criticisms alike) and state emphatically there is no ABC link, while making inaccurate statements.

Hmmm... "unsupported" is actually be equal to "lack of supporting". "Unsupported" presumes the interview studies are invalid, and the rat studies are not applicable. Neither of which is established. I do really like the statement: "preponderance of evidence against an abortion-breast cancer association." This shows the reliable sources interpretation of the scientific evidence, and does not give a POV value judgment on ABC evidence. I say POV in this instance, because Howe/Melbye/interview studies make "lack" and "unsupported" equally untrue. Whachathink? On got a call, going to stop typing now. Talk soon... man sometimes I wish Wikipedia had instant chat. - RoyBoy ⁸⁰⁰ 23:26, 2 October 2007 (UTC)

proposal

In light of the article, I propose the first paragraph be changed to include the following:

The abortion-breast cancer (ABC) hypothesis (also referred to by supporters as the abortion-breast cancer link) posits a causal relationship between induced abortion and an increased risk of developing breast cancer. At present, the hypothesis owes it's popularity to a concerted political effort by pro-life activists rather than a preponderance of scientific evidence. (Reference Mooney article) The proponents of the ABC theory postulate that in early pregnancy, levels of estrogen increase, leading to breast growth in preparation for lactation. The hypothesis proposes that if this process is interrupted by an abortion – before full differentiation in the third trimester – then more relatively vulnerable undifferentiated cells could be left than there were prior to the pregnancy, resulting in a greater potential risk of breast cancer.

--IronAngelAlice 00:00, 2 October 2007 (UTC)

I very much like the word preponderance, or something akin to that. I like "at present" too, as it doesn't assume the course of science on this subject. However, I am against the positioning you propose... essentially it is already stated in the third paragraph (see discussion above with Mastcell). I have something against prioritizing the politics of the issue, before actually defining what it is. My goal has been to reverse the popular (tabloid?) tendencies of focusing on extremes of the issue (pro-life vs pro-choice), and to put the science first, and the rest follows. As it stands now, I prefer your wording for the third paragraph sentence. - RoyBoy ⁸⁰⁰ 00:18, 2 October 2007 (UTC)

RoyBoy, I completely agree with you. Getting politics out of the top paragraph is generally a good idea. However, in a situation where the hypothesis is being popularized for political purposes, the politicization of the hypothesis should be acknowledged quickly. With this in mind, it should probably be explicit in the first or second paragraph that either 1) the hypothesis is currently being made popular by pro-life activists, or 2)that this hypothesis has been long-abandoned by the medical research community.

To the second end, we could change the paragraph to say:

The abortion-breast cancer (ABC) hypothesis (also referred to by supporters as the abortion-breast cancer link) posits a causal relationship between induced abortion and an increased risk of developing breast cancer. At present, there is broad agreement that the preponderance of scientific evidence shows no causal link between abortion and breast cancer. (Reference Mooney article, and WHO article) However, the proponents of the ABC theory postulate that in early pregnancy, levels of estrogen increase, leading to breast growth in preparation for lactation. The hypothesis proposes that if this process is interrupted by an abortion – before full differentiation in the third trimester – then more relatively vulnerable undifferentiated cells could be left than there were prior to the pregnancy, resulting in a greater potential risk of breast cancer. At this time, the hypothesis owes it's popularity to a concerted political effort by pro-life activists.(Reference Mooney article)

--IronAngelAlice 00:38, 2 October 2007 (UTC)

You're being reasonable, but #2 isn't accurate. Research is still being done on the ABC issue, though by all appearances the recent studies by Harvard etc. have reiterated the probable mistake of Melbye in haphazard statistical adjustment. (ie. accidentally removing ABC along with other multivariate/interacting confounding factors) By interacting I mean if you have an abortion, and there is an ABC link, then you are more sensitive to environmental carcinogens until you have a full term pregnancy. (ie. smoking, alcohol etc.) How scientists think they can remove these through statistical methodologies is quite curious from my perspective.

Editorial note: popularity should be prominence. - RoyBoy ⁸⁰⁰ 00:47, 2 October 2007 (UTC)

All statistical analysis is an approximation of the truth (though in this case, I think the Harvard study and others have made the best possible approximations). The underlying issue is that you can't prove a negative. But if study after study, using the best available methodology, fails to find a link, then most would conclude that no link exists. At least until presented with convincing evidence to the contrary - evidence which, at present, is non-existent. I think perhaps too much weight is being placed on the rat studies. It's completely unremarkable that an association in an animal model failed to translate to humans. If anything, it's the rule rather than the exception. MastCell ^Talk 03:02, 2 October 2007 (UTC)

Huh? I am not talking about generic "statistical analysis". The best methodology is the tried and true (time consuming and expensive, especially in larger studies) case-control matching (the standard and basis of epidemiological statistical analysis) which provides best relative risk results; they don't do that as much anymore, especially on large data sets. They are making educated guesses on how much breast cancer incidence (overall in a given population over time) is going up because of confounding factors. Anyone can find this out, by checking what region specific data they used to substantiate x, y and z confounding factors to be removed. Instead of doing case-control analysis they are making sweeping adjustments based on assumptions of "established confounding factors". When I see something like this, I take it to mean: "We have not established what the confounding factors were for this data set (for X reasons); so we will guesstimate based on generic studies and how much the overall incidence actually went up by."

The reliability of those studies and their methodologies is an assumption (based on what one expects to find), not a reality. So if they get a positive result, the assumption becomes that there was a confounding factor "missed" and needs to be "accounted for." I have not assumed, and I have attempted to understand how they back up their adjustments, thusfar I have been disappointed in what I find, or rather don't find within the studies. Don't believe otherwise please.

I mean come on Mastcell! Think about it. Scientists spend all this time and effort removing confounding factors; all of which involve hormones, without exception, every single breast cancer factor they have to account for involves hormones (even genetics); yet early pregnancy which makes a women glow and initiates significant changes to her body and breasts would have Zero impact? I'm not asking you to change your position, I want you to consider – seriously – that the scientific consensus is not based on bedrock evidence.

Without the rat studies there would be no ABC issue, you're both getting way off track. Also it is my understanding everything found to influence cancer in rats, has been found to do so in humans. Without exception. If I'm incorrect, please educate me on the efficacy of rat models. - RoyBoy ⁸⁰⁰ 20:45, 2 October 2007 (UTC)

I agree with Mastcell. It's clear that the top paragraph should be rewritten. --IronAngelAlice 16:08, 2 October 2007 (UTC)

Tweaked maybe, rewritten, doubtful. - RoyBoy ⁸⁰⁰ 20:45, 2 October 2007 (UTC)

Hmmmm Dont you think we could add something that I thought of in the first 5 seconds of reading this? I mean who cares if abortions increase the chance of breast cancer? Smoking does too but theres no laws against all smoking. Plus I say breast cancer is better than having a child you can't take care of... we have come a long way in medicine - not in welfare... —Preceding unsigned comment added by 198.138.41.69 (talk) 18:42, 7 October 2007 (UTC)

I think women would care about something effecting their health; also smoking is addictive, so it's not a precise analogy. Also, being informed about a ABC association can enable a women to make decisive decisions immediately that can reduce her risk. - RoyBoy ⁸⁰⁰ 04:12, 8 October 2007 (UTC)

Maybe best to use the subjunctive mood there, since no association has been shown to exist... :) MastCell ^Talk 04:49, 8 October 2007 (UTC)

Heh, well I could have added some qualifiers such as "might reduce" and "could effect"... but even if an ABC association is lacking supportive consensus; I don't think it's a stretch to say abortion, particularly later abortions, impact a women's health in some way. I'm going to tweak the lead; and eventually I'm also going to be adding the statistical analysis by Carroll along with this qualifier... which makes me think I'm going to tweak the sympathetic write up about Family Health International. - RoyBoy ⁸⁰⁰ 16:10, 8 October 2007 (UTC)

I'm pretty strongly opposed to using the Journal of American Physicians and Surgeons as a reliable source on medical/scientific matters (this has come up elsewhere). Aside from the well-documented fact that it is a political advocacy journal masquerading somewhat clumsily as a medical journal, it is not indexed by MEDLINE/PubMed nor Web of Science, which is a bare-minimum requirement for scientific journals in terms of credibility. I think JPandS is a reasonable source for what a specific slice of the American right wing believes, but not a reliable source on medical/scientific matters to be juxtaposed with the National Cancer Institute or American Cancer Society. MastCell ^Talk 16:43, 8 October 2007 (UTC)

Yup, found that out quickly enough through Wikipedia, that's why I put it at the very bottom, outside of the Scientific section. I wouldn't want to provide the journal with any undue weight. - RoyBoy ⁸⁰⁰ 18:30, 8 October 2007 (UTC)

I can't say I'm excited about it - I sometimes feel that JPandS is cited more often than the New England Journal of Medicine as a source on Wikipedia's medical articles - but at least your passage provides some context. MastCell ^Talk 19:56, 8 October 2007 (UTC)

A note

Due to confounding factors in real life, I don't have the time to give this article a full review, (Plus, I don't know how helpful it would be when i've already done it once, and therefore might be too familiar with the article) but there seems to be a rather odd oversight in the article nowadays. I look at the organization of it, and I see information on how the issue is politicized, the NCI workshop, a huge amount of space devoted to almost every study on the Abortion-breast cancer hypothesis, that weird court case, laws of the U.S. related to the article's subject, and a section for Carroll. However, there doesn't seem to be any actual material anymore in the body about what the ABC hypothesis actually is. The lead defines it, but the body never goes into what it is more specifically, thusly violating WP:LEAD by introducing unique content into the article, rather than being a summary. For example, some information about ABC that would be important would be when the hypothesis was first proposed, who proposed it, what sort of history was behind it before the 1980's if any, and things like that. Homestarmy 01:58, 30 October 2007 (UTC)

You're still immensely helpful, that will indeed need to be put in as History and/or Hypothesis sections. I would note however, every notable study (as far as I can ascertain from coverage) is mentioned here; there are dozens which are not. - RoyBoy ⁸⁰⁰ 02:54, 30 October 2007 (UTC)

Done drafts of both. Hope you like! - RoyBoy ⁸⁰⁰ 06:45, 4 November 2007 (UTC)

GA update

The article now meets the Good Article criteria, and will be listed. I implemented {{ArticleHistory}} for this article, and all previous peer reviews and GA reviews are linked through the template. The most recent GA review is here. Good work! Dr. Cash 07:16, 11 November 2007 (UTC)

Tremendous! Thank you very much for the review. - RoyBoy ⁸⁰⁰ 20:08, 11 November 2007 (UTC)

Tag/serious problems

I'm not asking for a reassessment because I've noted Royboy's work on and dedication to the article. However, the article has serious problems. A number of the studies lack context. While the article states 3 times how Brind alleges bias against the NCi, there is little discussion of other studies pointed criticisms of Brind specifically, nor is there any discussion of Brind, Lanfranchi and Malec's conflict of interest given Brinds much criticized Breast Cancer Prevention Institute (a front for pro-life propaganda) which was started by Brind and Lanfranchi and Malec's ABC. Both orgs are fronts for pro-life propaganda. They are unreliable, extremist groups and should be contextualized as such. The problem is that outside of Brind's studies, all the article's refs supporting the link come from Brind, Malec and Lanfranchi in various pro-life fronts. Probably because ABC doesn't get a lot of play in the peer-review circles anymore.

• While the article defines causal relationships, it does not adequately define associations/correlations and uses the three interchangeably, or at least without context.

• It misquotes the Russo studies in an unequivocal contradiction of the study’s results. (See refs 2-4)

• Check out ref. 37. An article from 1990 used as citation for Brind’s criticism of study published in 1997, wow – prophetic. (Or possible OR)

• 49 and 72 are biggies too.

Approx. 25% of the sources cited are unreliable, taken out of context, misquoted, misattributed, libelous or OR.

Questionable References: *2,3,4. OR, NPOV, Verifiability and Gross negligence. Article states “Drs. Russo & Russo...conducted a study in 1980 which found that rats who received abortions had a "similar or even higher incidence of benign lesions" and carcinomas than virgin rats of matching age.” Citation cites study. However, per Patricia Jansen, Ph. D, in her article “Breast cancer and the Politics of Abortion in the United States” published in the journal Medical History in 2005 (quoting the study directly): Russo and Russo observed that "abortion left the rats highly susceptible to developing cancer, but that "'the aborted rats were at the same risk as virgin animals treated with the carcinogen'" Over the next two decades, however, their findings would be cited repeatedly as evidence that pregnancy begins a process of breast change which, when stopped by abortion, put female rats (and thus humans) at greater risk of cancer than those who had never been pregnant.” She cites pp. 497-498 of the Russo study. http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1251638#fn16

• 9. (sources 2 assertions) Breast Cancer Prevention Institute (BCPI) fact Sheets, a self-promotional pro-life vehicle started by Dr. Brind (Pres.) and Lanfranchi (VP). Unreliable source for scientific crit. Should come from appropriate peer-review journal.

• 10. Dr Brind’s Review of the Melbye/Danish Report, hosted by www.abortiontruths.net, no author, no publication date, no publisher, unless we are to assume Brind wrote it for AT. I think this might be the response that Science refused to print. Unreliable source.

• 11. "Abortion and Breast cancer Collection-TP", a personal pro-life page, which also misquotes the Russo studies.

• 23. BCPInstitute, another Laundered Brind ref. Unreliable source.

• 26. Another BCPI ref. Unreliable source.

• 38. (sources 3 assertions) Brind study

• 37. Article from 1990 used as citation for Brind’s criticism of study published in 1997, wow – prophetic to boot.

Still cannot find this 1990 citation. - RoyBoy ⁸⁰⁰ 19:02, 25 November 2007 (UTC)

It's been removed already. Sorry, thought I crossed this out. Done. Phyesalis (talk) 22:44, 25 November 2007 (UTC)

• 39. “Abortion Breast Cancer/article_two” a letter from Brind for ABC, Karen Malec’s self-promotional pro-life org.

• 43. “Before You choose abortion” by Steve Frezza. The article says the cited info was published by AIM, Accuracy In Media, but the link of the source has no publication date, no AIM website, unless you count its homepage: “Human Life International Ireland: An Irish prolife/profamily organization” dedicated to “Our Lady of Guadalupe and Our Lady of Knock” discussing the sins of “homosexuality” and abortion. It is the property of the author, Scott Somerville Esq.. So now we’re citing lawyers who blog about abortion and cancer as valid scientific criticism?

• 46. (OR) article states that AIM alleges bias due to funding from FHI, citation links not to a source which would substantiate this claim, but to FHI’s index page.

No, the citation you removed is there to merely to define FHI; the bias source is given in the previous cite. I can simply move that cite down to the bottom to clarify. - RoyBoy ⁸⁰⁰ 23:51, 21 November 2007 (UTC)

Still not convinved that "Frezza" is a viable source. Needs a good argument, this is far lower than BCPI, ABC and Catholic Citizens.org on the reliable source scale. Phyesalis (talk) 06:40, 23 November 2007 (UTC)

Well, I've always considered AIM to be a reliable conservative source. I consider FAIR a reliable source. Or is the issue that it was published by the author, and/or its conclusions? The only reason I've continued to use it (founding it on web.archive) is because it was the only source at the time which examined the political/control groups issues other than Brind. I've always been eager to use other sources (even conservative ones) to not have to keep coming back to Brind. - RoyBoy ⁸⁰⁰ 01:00, 25 November 2007 (UTC)

• 48. (sources 3 assertions) Brind Study

• 49. cites the Royal College of Obstetricians and Gynaecologists in March 2000 published evidence-based guidelines on women requesting induced abortion. The review of the available evidence at the time was "inconclusive" regarding the ABC link. They also noted "Brind's paper had no methodological shortcomings and could not be disregarded." However the next paragraph in the source clearly states that the NCI and the ACOG concluded in 2003 that based on a rating (1) well established, there was no causal link between abortion and breast cancer.

• 55. Criticism of a reliable scientific study sourced from an editorial by Karen Malec from the website “Catholic Citizens.org”. Unreliable source.

• 56. Brind study

• 58. “Abortion and Breast Cancer: The Scientific Debate That Never Happened” Another letter in CatholicCitizens.org.

• 72. Article has incredibly controversial quote attributed to Daling, (which contradicts everything I’ve ever read about her in academic journals) supposedly from Daling herself in the LA Daily News. However, the source is Angela Lanfranchi’s opinion letter in “The Age”, an Australian newspaper, in which she’s supposedly quoting Daling. Daling's objection to her words and work being repeatedly misrepresented in pro-life propaganda is well documented. This borders on libel. Thoughts? Phyesalis 07:58, 15 November 2007 (UTC)

OK, struck the resolved items (many fixed by RB, thank you). Will restructure list in the next two days, as requested. I haven't struck the partisan sources because there are contextual issues that I'd like to look into/tweak. Phyesalis (talk) 06:40, 23 November 2007 (UTC)

Detailed work, many of the issues raised will have be addressed in some manner (on the weekend or thereafter). However, Russo and Russo do specify within their study(ies) P.I. (pregnancy interrupted) rats have equal, or even higher levels of carcinomas and this is reflected -- weakly -- in the early data; and more significantly in the later data (I will scan the page/chart in question which is not directly referenced in the article); as a quick stop gap measure to anyone who may try to jump the gun.

Daling, it was an important quote (and had been challenged previously) and I took the chance and called her at work (on the phone). She did not want to discuss the subject (as I believe she had discussed it more than enough for her liking), but she did confirm the latter part of the quote (which was all I was aware of at the time). This is OR on my part, but I simply could not gain access to the archives for the newspaper; I have also come this in other more reflective sources. I still have no reason to question its accuracy, let alone consider it libelous! While I understand your point of reference and pro-life propaganda context; it is a conspicuously partisan point in your thoughtful review of the article.

On that note, it is too easy to define any conservative source(s) as unreliable. I've demonstrated at least several pro-choice sources unreliable; yet they seem fine to most people who consider themselves impartial. It is this double standard which motivates me to make this article the best it can be (and essentially point out everyone could be wrong). As such, I look forward to working with you to get this article to FA status. (IMO it still meets GA status, as the "big" problems aren't actually big at first glance; with the exception of a possible 1990 vs 1997 continuity problem) I am not looking into now, since if I start I won't stop... and I need sleepy time. Your efforts are appreciated and will bare neutral fruit. 12:11 am, whew, even this took me long enough to write with all the typos, wrong words and such. Fatigue is a harsh mistress. - RoyBoy ⁸⁰⁰ 05:17, 16 November 2007 (UTC)

Additionally "similar or even higher incidence of benign lesions" has quotes, which means I took it directly from the Russo and Russo study. This wouldn't be the first thing I've found wanting in Dr. Jansen's excellent article. - RoyBoy ⁸⁰⁰ 05:20, 16 November 2007 (UTC)

Take your time. I've tried to express my good faith by not asking for a reassessment. The nature of my review is a reflection of the nature of the problems. It might come off as partisan because the errors seem to be of a partisan nature. I don't have a problem with the information being present (except for the Daling quote) but I do have a problem with the quality, or rather lack thereof, of the sources used to support them. In an effort to avoid partisan criticism, I haven't argued NPOV. I touched lightly on the possibility of NPOV (re: Russo) and on OR. If the info cannot be substantiated by reliable and verifiable sources, then I'd claim NPOV.

Thanks.

Do I understand you correctly that when you say the article doesn't have any big problems, you mean that you're okay with unreliable, partisan, and self-published sources as legitimate sources for a subject that has volumes of peer-reviewed material? ABC is a fringe theory, and has been internationally regarded as such since 2003. It can only be substantiated outside of Brind's studies in sources that do not meet WP's standards for WP:Reliability or WP:Verifiability. The article gives WP:Undue weight to Brind's position.

This is a partisan subject so partisan sources will be there; trying minimize them is fine, but they contain important counter-points and information; which while in the minority needs to be there for a complete article. Reliability isn't a straightforward objectoin; perhaps undue weight is given to Brind, but his prominence on the subject makes it necessary to cite his points which haven't been discredited.

Sorry to worsen the threading issues here, but... partisan sources are potentially useful here as indications of what particular partisans believe about the issue. They're not useful as scientific rebuttals of the National Cancer Institute or the American Cancer Society. We can of course use partisan sources - in fact, it would be hard to cover the political aspects of this issue without them, and the political issues are really the major ones since scientific opinion is pretty uniform. However, we have to be careful not to present this as a debate where both sides have equal scientific validity when that is clearly not the case, as demonstrated by reliable non-partisan expert bodies and sources. Additionally, once a scientific issue is considered more or less adequately settled, researchers move on. Given that there's a pretty clear consensus that abortion is not a risk factor for breast cancer, most researchers in the field have moved on to trying to identify actual risk factors and improve treatment, and spend correspondingly less time rebutting the handful of people perseverating about abortion. That doesn't mean, though, that every utterance of Brind's that has not been specifically rebutted in a medical journal is accurate or notable. MastCell ^Talk 21:47, 16 November 2007 (UTC)

The question of OR has been settled, at least regarding the Daling quote. You acknowledge that the quote is the product of OR that cannot substantiate the assertion as stated. Your diligence is admirable, but OR nonetheless. It doesn't matter if the statement is true, it matters whether or not one can cite it with reliable and verifiable sources. The assertion that the quote came from the LA Daily Times is potentially verifiable, but since no has verified it, it must be stricken. Additonally, as this is an exceptional claim, it requires multiple exceptional sources. A controversial statement attributed to Daling that can only be verified by an editorial letter from a pro-life ABC advocate who does not source the quote (no date, no context, essentially hearsay) is not an exceptional source. Given the documentation of Daling's vocal objections to pro-lifers misappropriating her words and work, the quote should be removed. Because it is the product of OR, it should be removed immediately. Phyesalis 08:11, 16 November 2007 (UTC)

I should be able to verify it; till then, fine. I still question your metric for reliability of the pro-life source. - RoyBoy ⁸⁰⁰ 21:27, 16 November 2007 (UTC)

Royboy, perhaps when you verify it, you could get the paragraph that it's in and the paragraph before and after it. It would be a great addition to the history of the document and could be used as reference for future challenges. If we can get the name of the article, pg number, author, date and three paragraphs as context of the quote, I would accept it as an exceptional source and defend its inclusion in the article. How does that sound?

Mastcell, you make valid points. After rereading WP:Fringe theories, I don't object to the inclusion of extremist/partisan/unreliable sources (BCPI-Brind and Lanfranchi, ABC-Malec and Catholic Citizens of Illinois as a mouth piece for ABC-which would otherwise be an unreliable source of scientific criticism), per se. They should be contextualized as such. ABC is primarily the work of Brind, Lanfranchi and Malec. They aren't "critics", they are three particular and isolated people. I think the article needs to be rewritten/restructed to neutrally and accurately reflect this. The use of the general "critics" implies a broad base of work, all ABC content (as opposed to general research examining the possibilty of a correlative relationship - not causal) can be attributed directly to one of the three and should be.

Regarding a rubric for extremist partisan sources, I suggest that there are several levels of sources, some of which can be excluded. Home pages or personal domains, indiscriminant repositories of data compiled by the general public, websites that are clearly amateur for either pro-chice/pro-life prop. or amateur religious sites for the purpose of espousing pro-life prop. (we could collectively map out what this means but I think we all get the general idea), lists of data from unreliable sources, unsigned info from unreliable sources, and sources used as OR should be excluded. This does not include Catholiccitizens.org, while religious is not clearly amateur nor clearly for the purpose of pro-life prop. But CCI sources should be contextualized as partisan and appropriate refs should be attributed to Malec. Same with Brind, Lanfranchi and BCPI. Thoughts? Phyesalis (talk) 00:12, 17 November 2007 (UTC)

Rats and Russo

From "Rats": "Drs. Russo & Russo from the Fox Chase Cancer Center in Philadelphia conducted a study in 1980 which found that rats who received abortions had a "similar or even higher incidence of benign lesions" and carcinomas than virgin rats of matching age.[2]"

AFter having posted a cited reference to a contrary interpretation, I looked up the paper. In good faith, I will assume that the editors responsible for the "Rats" subsection honestly missed the fact that the quoted section, which comes from p. 506, refers only to benign lesions, not carcinoma. Nowhere in the paragraph, or page, does it say that rats who had "pregnancy interruption". Which would be weird if it did, since it clearly states in the summary at the beginning of the study:

"Therefore, while pregnancy and lactation protected the mammary gland from developing carcinomas and benign lesions by induction of full differentiation, pregnancy interruption did not elicit sufficient differentiation in the gland to be protective, and these animals were at the same risk as virgin animals" (p.497)

What this means is that while PI did not convey the benefits of pregnancy and lactation, it did not raise the risk of cancer. Now, Royboy asserts "However, Russo and Russo do specify within their study(ies) P.I. (pregnancy interrupted) rats have equal, or even higher levels of carcinomas and this is reflected -- weakly -- in the early data; and more significantly in the later data (I will scan the page/chart in question which is not directly referenced in the article); as a quick stop gap measure to anyone who may try to jump the gun." Of course the rats have equal levels of carcinoma with virgin rats, virgin rates are the baseline. It would be really amazing if PI rats had lower levels! But the paper clearly states that PI rats had equal risk rates for carcinogenisis as virgin rats. Perhaps you are confusing benign lesions with forms of cancer? The quote used in the sub-section is part of an assessment of lactation as a protective measure against HANs and cysts, having nothing to do with a link between abortion and cancer.

Don't worry about scanning charts (any attempt to re-interpret data contrary to the authors' assessment is OR anyway). I'll just change the section to accurately reflect the source. If, as you suggest, later studies do contradict earlier findings, perhaps you should source them instead of the study that clearly states the opposite. In light of this apparent misunderstanding, I think page numbers would be helpful. Phyesalis 10:43, 16 November 2007 (UTC)

I based that sentence on Table 2 Pg. 502 on "Incidence of Adenocarcinomas"; where P.I. has 77.7% incidence and Group V control was 71.4%. Are these percentages discussing purely benign tumors, if so then that is my misunderstanding. I have only looked into the precise definition of adenocarcinomas right now; if I understand correctly their initial formation is always benign? Russo and Russo stipulate the rates are "essentially the same" (pg. 503); rather than "the same risk as virgin animals" from Jansen. I would suggest that while my sentence fragments was inaccurate, it is more accurate (given the data), than Jansen's editorial of Russo's cautious interpretation.

The results aren't "clearly" anything, and I was quite antagonized at the suggestion I re-interpreted the data. However, I gave myself time to respond and had a great weekend involving Blade Runner (Final Cut) screening in Toronto; and then took my time to go over the study and figure things out. I did re-interpreted the meaning of the sentence; but not the data. If I had done that... that would very disappointing. - RoyBoy ⁸⁰⁰ 00:56, 20 November 2007 (UTC)

Additionally there was a late trip to an airport, so I am quite tired and impatient right now, so I will only make modifications I feel are necessary. I am following your example and will begin separating issues into sub-headers; I have graduated Russo up; as to avoid nesting headers unnecessarily. - RoyBoy ⁸⁰⁰ 01:25, 20 November 2007 (UTC)

Cool (Blade Runner - yay!), I too had to step back when I first started dealing with the Russo info. My point is that when an author has clearly stated a conclusion or findings, as the Russos did in their summary (First page, and correctly quoted by Jansen: "Therefore, while pregnancy and lactation protected the mammary gland from developing carcinomas and benign lesions by induction of full differentiation, pregnancy interruption did not elicit sufficient differentiation in the gland to be protective and 'these animals were at the same risk as virgin animals [bolded material is the Jansen quote taken word for word form the Russo study]) any attempt to use data to suggest otherwise is OR. When dealing with scientific studies, one should use the author's interpretation, not look at individual charts etc, to try and discern what the author has already clearly stated. "Interpreting" data in a manner inconsistent with author's conclusions is OR, per WP:OR "Primary, Secondary, and Tertiary Sources".

It wasn't my intention to antagonize you, quite the opposite. I think I showed restraint. You can see why, when the authors clearly state in the abstract and the study that they are the same (or "essentially" the same - enough so that they felt comfortable stating that they were the same in the summary), and you look at one chart and determine otherwise, why I might have thought that the work was either cribbed from an unreliable site or the product of OR. The fact that you have voiced objections to the work of a peer-reviewed academic (Jansen), and were unable to read the study in question and determine that Jansen was indeed correct (because she had quoted them word for word), could also lead one to uncomfortable questions. I did not suggest that you were engaged in OR. I assumed good faith and asked if "editors" had missed the appropriate sections and misunderstood what benign lesions were (I didn't know who had originally added the info and didn't think it was important). I did mention OR in brackets as part of a sentence directed at you because you said you'd go back over the charts/tables - if you had done that, instead of relying on the author's interpretation, that would have been OR. You acknowledge that you "did re-interpret the meaning of the sentence", and I'm not suggesting that you are engaged in OR. No harm, no foul? (If not, please feel free to continue this on my talk page.)

And thank you for improving on my change by moving the Russo info up. I didn't want to do too much without giving y'all a chance to weigh in. Phyesalis (talk) 05:51, 20 November 2007 (UTC)

I've improved it further by removing your OR. Refuted? Uh, no... Melbye refers to the Russo and Russo hypothesis of abortion effecting cell cycle length as a possible explanation for their sub-group positive results. And rats are mammals, and hence have breasts. - RoyBoy ⁸⁰⁰ 22:19, 20 November 2007 (UTC)

Unfortunately had to revert a lot of changes to Proposed mechanism. You made redundant statements; and strayed (slightly) from Michaels statement (not research) which stands quite nicely on its own. The preamble that was added did not clarify early terminations is what Michaels is discussing; so this does not address/effect later terminations. - RoyBoy ⁸⁰⁰ 02:17, 21 November 2007 (UTC)

Umm, would you mind providing reliable proof for your assertion that rats don't have breasts - Dr. Mor (head of which epidemiology org?) states in the article *explicitly* that they don't. Mammary glands are not breasts, as breasts are what contain mammary glands - thus it is possible to have mammary gland tissue while not having breasts. Cows and goats have udders, not breasts, whereas most other animals have mammary glands or mammary tissue. The Russo study, if I recall correctly, never mentions rat "breats" or "breast tissue" but "mammary tissue". I'm going to change it back. Until you can find a source which contradicts Mor, it should stay. Even then, because it is sourced, you can only contradict it, not remove it. I'm going to review your changes. I will probably revert back to my original work and reintroduce your changes there. Phyesalis (talk) 02:37, 21 November 2007 (UTC)

Mammary glands are similar tissue whether in breast, udders or teets, but point taken. Been a long time since I read that section. I reverted the breasts... heh, rather a weird thing to say. - RoyBoy ⁸⁰⁰ 04:41, 21 November 2007 (UTC)

After reviewing changes - I belive the wording in Mastcell's last edit is more accurate, per consensus on refutation of theory and fringe position of supporters. The intro should not be a line by line summary of the article. The intro is too long. Criticism of international studies belongs in the body of the article not the intro - this undue weight given to fringe advocates. Phyesalis (talk) 02:46, 21 November 2007 (UTC)

While I don't completely agree, Brind is undoubtedly fringe (which was why he was removed from an old version of the lead in the first place). I get a vague sense Brind was named in order to push criticism out; for now I can't muster sufficient reason them in the lead; however given his prominence on this issue... and the importance of the interview/response bias meme... I'm thinking that (tied in with Brind), should stay. How's that edit? - RoyBoy ⁸⁰⁰ 04:32, 21 November 2007 (UTC)

First, thanks for reverting the breasts (lol). I'm going to take a crack at the intro, as well as change some wording - the Jasen (!!! - I've been putting an extra 'n' in her name all this time - got to fix that) bit needs to be changed - her article does not discuss "preliminary findings" but the refutation of the link. Mind taking a look over when I'm done (maybe 10 minutes?) and let me know. Phyesalis (talk) 05:39, 21 November 2007 (UTC)

Tweaked "no increase" to "similar risk" to better follow their results. - RoyBoy ⁸⁰⁰ 22:17, 9 December 2007 (UTC)

rejected

Loaded term (partially because it is ambiguous) that implies ABC hypothesis has no science behind. Quite the opposite, studies such as Melbye and Michels indicate high risk sub-groups; and Howe indicates there questions open for consideration; and Daling results speak for themselves regardless of your interpretation(s) of her outlook on the ABC hypothesis. Please do not overreach. I have removed it from the lead; also your championed sentence is accurate, yet very misleading... giving the impression only a pro-life person would consider the ABC hypothesis plausible/real.

Situating a polarized issue with polarized conceptions isn't helpful. Is it NPOV? I'm unsure, as I take great care in leveraging policy when it comes to issues of tone. - RoyBoy ⁸⁰⁰ 01:25, 20 November 2007 (UTC)

I'm sorry, I didn't see this note before I reverted. I did leave a note regarding the "rejected"/"unsupported" issue at the bottom of the page - can we move this discussion down there? I don't think it's "polarizing" to describe the scientific community's take on the subject (as sourced to the NCI, American Cancer Society, etc). Certainly, reasonable people ca disagree with the scientific community's view of the topic, but that doesn't change the fact that their view is what it is. It has been rejected, verifiably so. That doesn't mean it has no supporting evidence - it would never have gotten as far as an NCI workshop, nor would this many studies have investigated it, if there were not some basic-science or animal model data to support it. That doesn't change the fact, though, that the hypothesis as applied to humans has been rejected. MastCell ^Talk 01:37, 20 November 2007 (UTC)

Certainly reasonable people can disagree, but how can we truly know a scientific communities view on a politically charged fringe subject; where there is little incentive to dissent? Or in short, we can't know what is a reasonable median. (I think I do, but it involves trying to read between the lines of pro-life sources and pick out the accurate parts, certainly an unscientific endeavor.) But the scientific consensus on the ABC issue is unscientific; by any stretch of the imagination there would be doubt stemming from deviations this article points out. But there isn't, the conservative "there might be exceptions" interpretation goes out the door when it comes to rendering the evidence for public discourse. We can, and I submit, we must do better. - RoyBoy ⁸⁰⁰ 05:32, 20 November 2007 (UTC)

In real life, yes, I agree with you. For Wikipedia's purposes, though, you seem to be advocating taking verifiable, well-sourced information and parsing based on our personal level of faith in the scientific consensus. Anyhow, scientific consensus isn't set in stone, it's hardly ever 100% unanimous, and it's subject to change should additional convincing data be presented. But right now, there can be no doubt that there is a scientific consensus that abortion does not cause breast cancer. If that changes tomorrow, then I'll be the first (or second, maybe) to update Wikipedia to reflect the change, but we can't anticipate a change in scientific consensus based on our personal reading of the evidence or valuation of the NCI's expert opinion. Not on Wikipedia, anyway. MastCell ^Talk 18:59, 20 November 2007 (UTC)

LOL, second; nicely played buddy! I think I just want to remind everyone that NPOV isn't just about the public discourse/consensus, but of the scientific nitty gritty; and to not shy away from having both in the article... even if they are, at times, incongruous. There seems to be a notion among some that we are required to tweak the article to present consensus tone at all times. - RoyBoy ⁸⁰⁰ 02:28, 21 November 2007 (UTC)

Patrick Carroll

Section should be reverted. You have misappropriated policy, don't do that. WP:Crystal ball refers to future events; as in if the section talked about the study happening soon. It does not refer to predictions of sources. Even if policy were tweaked to include this instance (it shouldn't be, since many notable writers make predictions that are used by Wikipedia) you are missing the forrest for the trees. It is crucial this pro-life reference, which is already being used in the public sphere; be mentioned here and given appropriate context.

To not do that makes for an incomplete article. Certainly we could argue the notability of such a non-notable / unscientific source, however, I think its notability was given appropriate weight as a small sub-section in the political section at the basement of the article.

There is more to cover, but I need to be outta here. Much more to come from us both, I'm sure. - RoyBoy ⁸⁰⁰ 01:25, 20 November 2007 (UTC)

Agree with RoyBoy here. MastCell ^Talk 01:39, 20 November 2007 (UTC)

I didn't purposely "misappropriate" the policy. After re-reading policy, I see my mistake. Thanks for pointing that out! I have a better understanding now. Objection withdrawn. It seems we all have a lot on our plates. I'm happy to continue this civil and productive pace. Looking forward to your input. Phyesalis (talk) 04:02, 20 November 2007 (UTC)

Can anyone say whether the statistical study by Carroll has been challenged by other statisticians? The only attack I saw was from an (unqualified) Horizon editor. ^rossnixon 01:18, 28 November 2007 (UTC)

Rossnixon, I've followed your note here. Unfortunately Carroll is not a scientist, he is an actuary, not an epidemiologist. Even if he were a legitimate scientist - one study in a pro-life journal does not topple the National Cancer Institute, the Netherlands Cancer Institute, the American College of Obstetricians and Gynecologists, the Royal College of Obstetricians and Gynecologists, New England Journal of Medicine, and the head of Epidemiology for the American Cancer society. Also, his 2001 study was labelled as "mischief-making". And this study hasn't been reported much in the mainstream media (you should read all the pro-life rags bitterly complaining about it), because he isn't a legitmate epidemiologist. Phyesalis (talk) 03:11, 28 November 2007 (UTC)

So, his statistics have not been challenged? Thanks. ^rossnixon 01:19, 29 November 2007 (UTC)

I concur with the edit, but not the argument used in the summary. "hard for abortion to be the "best predictor of breast cancer" if there is no recognized association." That isn't what the evidence says; it stipulates no causal association. That is not the same as "no association", also it is a "best predictor" only among the seven factors it choose to examine. Though I find it very strange fertility (many children) wasn't the best predictor... unless they had very few in their dataset. - RoyBoy ⁸⁰⁰ 05:07, 4 December 2007 (UTC)

Relative risk

CI must not include (1.0-1.9)? It is my understanding if the CI does not include 1.0 that makes it significant; as Howe's abstract indicates. - RoyBoy ⁸⁰⁰ 22:55, 2 December 2007 (UTC)

Also, must not isn't the kind of language I'd expect from scientific explanations. - RoyBoy ⁸⁰⁰ 23:49, 2 December 2007 (UTC)

Article is now corrected: RR with CI not including 1 is significant.Ekem 01:33, 4 December 2007 (UTC)

Apparently there is an alternative opinion on that, which I'm curious to learn the origins of. - RoyBoy ⁸⁰⁰ 05:08, 4 December 2007 (UTC)

I asked for Ekem's input by asking for an opinion on the Dr's Mess (because I had a sneaky suspicion that we were both wrong). Turns out, I think we were. I knew that confidence intervals that went from <1 - 1 or >1 were statistically insignificant. That plus this from Jasen's article:

For example, a relative risk of 1.5 suggests that women who have had abortions have a 50 per cent higher risk than other women, when all other factors have been controlled. The meaningfulness of such figures depends on many elements, including the ability of researchers to separate abortion from confounds and from other risk factors, the appropriateness of the control group, the numbers of women in the study, the accuracy of reporting abortions, and the extent to which low levels of relative risk (below RR=2.0) can be considered significant.

So between the confidence intervals and Jasen, all the studies appeared to be statistically insignificant.

My initial point was the article's previous treatment of confidence intervals was confusing. Consider this diff [3]. It basically states that a CI of 0.9 to 1.04 is statistically significant. My error was in an edit that conflated the two (the figure of 1.1 - 1.9). I'm in the process of bringing the studies in line with Ekem's edit. Phyesalis 06:20, 4 December 2007 (UTC)

Going over the history, I realize that I came off a bit cranky (I think the "rejected" debate was once again rearing its ugly head) when I wrote the summary edit. I apologize. I just want to get this straightened out. Clearly the issue of the significance of these figures is a contested issue (I just re-read Jasen again - Royboy, would you mind taking another look at it and give me your take on her assessment of RR. It's in the first or second section after the intro.) Phyesalis 06:28, 4 December 2007 (UTC)

I'm very pleased things are coming together as collaboration is increasing and nastiness is decreasing. Even with my authoritarian note above below about Meta-analysis, you took the high road and just discussed the issues at hand. We've seriously turned a corner; and I'm feeling the Wikilove. Yay! Certainly, Jasen... I think I'm very overdue to re-read IMO the best ABC write up in existence. Beforehand though I'm going to relax, eat and bask in this moment for a bit. I'm in a good mood, where's my milk chocolate covered almonds? Yum! - RoyBoy ⁸⁰⁰ 23:28, 4 December 2007 (UTC)

Yay! I've had to step back and enjoy the moment as well. I'm going to keep the glow and come back later to address your other comments... Phyesalis (talk) 02:06, 5 December 2007 (UTC)

Re-read Dr. Jasen, from what you have quoted and surrounding text including:

The interpretation of statistical findings was inconsistent as well, for a certain level of risk (usually in the range of RR=1.2 to 1.5) was considered significant by some and not by others—an issue still hotly debated in recent years, and part of the reason why some of the same studies have been used as supporting evidence by opposing sides in the debate.

How do we reflect this in the light of unaccounted confounds (eg. Howe and genetic history) along with varied opinions of significance. The easiest path would be to reference the meme(s) Jasen presents in the Background RR explanation as a sort of disclaimer; but to maintain within study sections results are significant based on CI, and of course keep criticisms intact as possible explanation(s) for said results. I'll attempt a draft here on talk. - RoyBoy ⁸⁰⁰ 22:11, 9 December 2007 (UTC)

The majority of the results in epidemiology are calculated as a relative risk (RR) where 1.0 is no risk, but results above like 1.21 is a 21% increased risk and results below such as 0.8 is a 20% decreased risk. Relative risks are not necessarily significant. To help assess this a RR is followed by a confidence interval in brackets that shows the likelihood (with 95% confidence) that the RR is of significance. Any RR with a confidence interval that does not include a value of 1 could be considered significant. For example, the confidence intervals (0.3 - 0.9) and (1.5 - 7.8) are statistically significant, whereas the confidence intervals (0.89 - 7.34) or (0.5 - 1.1) are not. With more data the confidence interval becomes smaller; making it an indicator of the result's statistical reliability.

When a RR result actually becomes significant is a difficult and contentious issue. As a small result of 1.41 (1.1 - 1.6) even with a significant confidence interval may be inaccurate because of recall bias, incomplete data, missed confounding factors, imprecise controls or statistical analysis. These possible flaws in any study can effect its result. If that result is close to 1.0 it is probable that correcting those problems may change the confidence interval and the results significance.

The number of (X/Y breast cancer cases/controls) gives X as women in the study who have had induced abortion(s) and Y is women with no abortion and miscarriage history. This dataset is used when calculating the RR and provides a way to compare the size of one study to another.

Tweaked 1st and 3rd paragraph, while the 2nd paragraph is new and attempt to address concerns Jasen raises on RR significance. I removed the CN note on the X/Y note. This isn't verifiable, as it is simply noting what the numbers given within the article mean. It is a self-reference. - RoyBoy ⁸⁰⁰ 00:51, 10 December 2007 (UTC)

Meta-analysis

"This study has been criticized for methodological flaws and the blurring of association and causation." I removed the line as the first part is redundant and vague as flaws are already specified; as to blurring if a source can be found (or is already present) please do add it to the first paragraph of the section.

The main issue I have with it though is that its in the "Hypothesis and proponents" section. While Brind was the lead author, there were other scientists involved who are pro-choice and would likely take exception to being labeled "proponents". It should be back where it belongs, in the meta-analysis section under "Epidemiological studies" with Beral; they are both meta-analyzes. If Beral wants to use different terms to define a different methodology they undertook, fine, but they looked at 50+ case-control studies... that's a meta-analysis of existing data. - RoyBoy ⁸⁰⁰ 00:08, 3 December 2007 (UTC)

There was one, Chinchilli. And I can add him in, including the fact that he wasn't as exuberant as the three pro-lifers. However, Chinchilli isn't actually a proponent of ABC. He's the pro-choice stats guy they hired to give the study an air of legitimacy. He only did what, one or two studies with Brind. So I think that info about Chinchilli is appropriate in the Meta-analysis section. Phyesalis 05:01, 3 December 2007 (UTC)

Excellent read on Chinchilli, some good memes there that you can add as you move the study back to where it belongs. The only practical purpose it serves there is to fill out a short section. Apart from that it's your unnecessary separation (as it contains Brind's name). Move it back please, and get rid of your continued OR determination of significance and "other studies". It's transparently incorrect, partisan, unreferenced to boot and wasting our time. - RoyBoy ⁸⁰⁰ 04:35, 4 December 2007 (UTC)

RoyBoy, you're getting that "ownership" tone again, very authoritarian. I'm not sure it's polite to tell people what to do. I will address some of the issues tonight in a series of edits. Besides, there is no "meta-analysis" section now that we've identified the Beral study correctly as the author does. Additionally, as Brind et al. is the only study to assert causation, and the only study that alleges the ABC theory of causation, I think it should come first in order to represent the theory. It would be disingenuous to categorize the other studies as if they had anything to do with causation. They didn't study causation, they studied association (because you first have to prove association). Big difference. Phyesalis 05:06, 4 December 2007 (UTC)

You're absolutely right, as I was messing around with the Blade Runner article just now; I once again let my annoyance at things I disagreed with take precedent over discussing the issues. I apologize, and will make a conscious effort to not do that. However, my authoritarian (certainty) tone stems from my assuredness in "something" not being right; which is simply... disconcerting, but I should know things don't need to get fixed right away. After all, my mistakes/shortcomings have been in the article for years! Sorry for that and the tone. Though from a psychoanalytical perspective I was projecting the certainty you appear to have about what is and is not significant. - RoyBoy ⁸⁰⁰ 23:10, 4 December 2007 (UTC)

What I should have said instead is, look at Joel Brind and the SkepticFiles.org reference I found. They are no fans of Brind, but they're article references medical researchers who point out Brind's own study is "barely statistically significant". I feel that's accurate, succinct and ideal weight. - RoyBoy ⁸⁰⁰ 00:31, 5 December 2007 (UTC)

Thank you, I think we've both managed to take this in stride rather well.

As for SkepticFiles.org, it's from 1996. Try to find something post-NCI, and maybe a little more reliable than skeptics.org. If you can, I see no reason why it couldn't be included as a counterpoint to the well-documented criticisms. Phyesalis (talk) 01:52, 5 December 2007 (UTC)

Um... actually it is also a criticism of Brind and his study, showing significance might be in the eye of the beholder. So I've found it to be a useful clarification (not a counter point) to describing Brind's results. - RoyBoy ⁸⁰⁰ 01:50, 10 December 2007 (UTC)

Sorry, I didn't understand what you were trying to do with it. Didn't mean to piss in your Cheerios. I agree that the proposed wording is a good way to describe Brind's results. I should have said "Good work!" But my point is that it's not a good source. It might be useful but it's not reliable (home website doesn't mention who runs it and I am not familiar with its reputation). The cited 1996 listserv source does note an editor but I don't feel comfortable using it, particularly since I've been adamant about not using unreliable pro-ABC sources. However, since I don't think the info cited is controversial (we both agree that it is useful wording), I'd be open to an argument for its inclusion if you still want to keep it. Phyesalis (talk) 06:04, 10 December 2007 (UTC)

Hehe, well adding it would be okay, but to be honest I was initially pushing for it so that I could just get a compromise and remove insignificant. You know, the good ol' overshoot to get what you want. But it could be added along with a ref to Jasen (who would substitute as a medical expert), to help solidify/clarify that small RR's are questionable. That may work even better. - RoyBoy ⁸⁰⁰ 22:55, 10 December 2007 (UTC)

Why say JAPS is "not a medical journal"

I think I understand, that "medical journal" connotes something very specific, but at it stands, the statement looks like OR. I think that needs to be remedied.LCP 16:31, 15 November 2007 (UTC)

Saying that it's not a medical journal, while essentially accurate, is probably unsourceable. Many of the authors of articles published in the journal are lawyers or other non-medical personnel. The journal and its publisher have a primarily political rather than medical agenda. Numerous mainstream sources have described the journal as scientifically suspect and ideologically driven. Most importantly, perhaps, the journal is not indexed by MEDLINE or PubMed, which means its articles do not turn up in standard literature searches and thus are not cited by other authors and have essentially no impact on scientific discourse. But I do agree that a blanket statement that it's "not a medical journal" is probably unsourceable. I'm fine with leaving it as "controversial and politically conservative", which I think is much easier to source. MastCell ^Talk 17:30, 15 November 2007 (UTC)

Take a look - I've reworked the short paragraph and journal description to be a little more precise and stick closely to available sources. MastCell ^Talk 17:41, 15 November 2007 (UTC)

I am concerned that the way you put it invalidates Carroll’s analysis. And in spite of the politics of the journal, I am not sure that Carroll’s work deserves to be written off. On the other hand, I don’t want to give Carroll more credence than he deserves. At this point, I’ll defer to other editors as I am mostly out of my depth. I am not a statistician. However, I would prefer something like this:

Patrick S. Carroll used statistical analysis to forecasts, for the year 2025, higher breast cancer rates for Czech Republic, England and Sweden and lower for Finland and Denmark based on abortion trends. He published his article in the peer-reviewed Journal of American Physicians and Surgeons, a politically conservative and controversial journal with an explicit pro-life stance.

LCP 18:02, 15 November 2007 (UTC)

I think that it becomes an issue of undue weight. The fact is that Carroll's findings, published in a journal like JPandS, have been accorded little or no weight by the scientific community. It would be inaccurate to give the impression that they are considered as scientifically valid as the other studies cited. The Guardian wrote of this particular study that ... it comes as little surprise to discover the Papri study was funded by the anti abortion group Life, and that the rightwing Journal of American Physicians and Surgeons, where it was published, is home to arguments such as: "The gay male lifestyle shortens life expectancy by about 20 years." The BBC, discussing an earlier study by Carroll and funded by a pro-life activist group, quotes the Royal College of Obstetricians and Gynaecologists as saying that the study authors were "mischief-making" and that there is no good evidence of a link. Perhaps the Guardian source should be included with regard to this particular study. Regardless, I think the study's funding by a pro-life activist organization is noteworthy, particularly as its findings differ so dramatically from those of more weighty studies. MastCell ^Talk 19:56, 15 November 2007 (UTC)

My source is from a letter to the editor in my local paper by the Head of Surgical Oncology at McMaster University. I have not come across anything to indicate (apart from its name) that JPandS is an actual medical journal associated with any accredited medical group or association. I did not put the source, as it was recent, and I simply didn't think it was online yet. Also, it is inferred any reliable medical source would be listed with PubMed or some other medical journal archive. (just as any accredited University would archive their doctoral theses at the Library of Congress) - RoyBoy ⁸⁰⁰ 05:29, 16 November 2007 (UTC)

Yes, and I'd add that even quite a few unreliable (or at least fringe-y) medical journals are indexed by MEDLINE - the bar isn't that high and indexing is not a de facto sign of quality. However, the absence of MEDLINE indexing is a pretty good sign that a journal carries little or no scientific weight. MastCell ^Talk 18:54, 16 November 2007 (UTC)

Rejected vs. unsupported

I think "rejected" is an appropriate term, and is supported by our sources (e.g. the NCI workshop) - this hypothesis really has been rejected as inaccurate by the scientific community. If new evidence turns up, then of course it will be re-evaluated, but I'm not sure why we need to dance around its current status, which is well-documented. Particularly as this article already takes a somewhat more sympathetic POV toward the subject that scientific sources do (which is fine, this isn't SPOV after all - but still we should accurately characterize the scientific community's view here). This phrasing is problematic in any case - the hypothesis is not unsupported by the scientific community - it's unsupported by the scientific data (in the judgement of the scientific community). MastCell ^Talk 01:32, 20 November 2007 (UTC)

I agree with MastCell. This has been "settled" science since 2003. Phyesalis (talk) 04:24, 20 November 2007 (UTC)

Nope. That's like saying "prions" not causing neurological disorders had been settle in the 80's. Scientific consensus is what it is, but it isn't a sound yes/no rejected/confirmed; or even Busted/Plausible/Confirmed metric of Mythbusters. It is more nuanced that than that, the NCI be damned quite frankly; because they cannot overturn years/decades (depending on your understanding/position on response bias) conflicting evidence on a dime. Their workshop was an editorial response to Bush administration changes to their website. They didn't produce anything, they merely did a quick meta-analysis on the studies which were "considered" (unscientifically I might add) valid. The weight given to them seems wrong to me.

Okay, re-read your paragraph again; yes, that is actually a good read of the situation. But I would stress the NCI's standard of what constitutes "valid scientific data" is politically/sociologically influenced. They are not robots operating independently of reality; and they sure as hell didn't bring new data to the issue. "Settled" my cute behind (opinions vary), pre-term pregnancy having indications of increased breast cancer risk isn't a lark; it is consistent with the ABC hypothesis. Russo's studies consistently indicate a weak ABC correlation. Does that translate to humans? Well maybe not, as I pointed out in the mechanism section... but maybe it does in certain circumstances. Hormones are certainly not always beneficial. - RoyBoy ⁸⁰⁰ 05:07, 20 November 2007 (UTC)

I think you've hit the crux of the matter. Prions were settled in the 1980's, and if Wikipedia had existed, I would expect it to have said that prions were harmless. That's Wikipedia's role: to summarize what other reliable sources have said, not to be ahead of the curve and out-analyze them. Then, when the scientific consensus changed to incorporate prions as a cause of disease - then Wikipedia would reflect that new understanding. Scientific understanding may change - in fact, it definitely will. Wikipedia, though, should lag suitably behind changes in the scientific consensus rather than trying to anticipate them. "Rejected by the scientific community" is just a statement of fact. Tomorrow it could be un-rejected - if the WHO comes out with a statement that suddenly supports ABC, then we'll update the page. But for today, it's been rejected. The NCI didn't "overturn" anything - they summarized the available primary data, as expert panels do, into a reliable secondary source which, frankly, we should be giving far more weight than the article currently does (and far more weight than Brind's hardly impartial opinion). I respect that you've come to a particular conclusion, but Wikipedia's take on this needs to be based on a proportional representation of what experts have to say on the matter, in the form of secondary sources like the NCI workshop. MastCell ^Talk 05:19, 20 November 2007 (UTC)

When I sober up, I hope I still disagree with that. Prions and Charles Van Doren shall be my basis. - RoyBoy ⁸⁰⁰ 05:36, 20 November 2007 (UTC)

RoyBoy, you must be on a roll, because you also hit upon another issue - ABC posits a causal relaitonship, not a mere correlation. Regardless of the fact that some weak evidence for a correlation has been found in some studies (different from both an established correlation and a causal relationship), no major organization recognizes a correlation, never mind recognizing a causal relationship. And I agree wholeheartedly with Mastcell that this article gives undue weight to Brind's fringe position. Phyesalis (talk) 06:03, 20 November 2007 (UTC)

Indeed, in the last GA review a while back; I made a point of tweaking the lead to say there is "potential" breast-cancer risk, as the risk increase would come from damage over time; not the ABC mechanism itself. I do not believe ABC causes breast cancer, but is rather a co-causal factor; or perhaps more accurately indirect aggravating factor. Actually come to think of it, I am unsure ABC ever posits a "causal relationship", some pro-lifers certainly do; but do the main advocates/champions make that error? - RoyBoy ⁸⁰⁰ 22:02, 20 November 2007 (UTC)

The main advocates are pro-lifers - certainly the hypothesis has little or no remaining traction in the scientific community. Certainly every "advocate" cited here posits a causal relationship. MastCell ^Talk 22:21, 20 November 2007 (UTC)

I'm not so certain, at least from looking at the conclusion of Brind's meta-analysis; which stipulates a "significant positive association", and the abstract states "adversely influence a woman's subsequent risk of breast cancer". I do not spot causative verbiage immediately. - RoyBoy ⁸⁰⁰ 22:32, 20 November 2007 (UTC)

Brind's stance is well-documented, and "adversely influencing" the risk of breast cancer is a clear assertion of causality. If this was solely an association, then it would not "influence" the risk of breast cancer. MastCell ^Talk 22:57, 20 November 2007 (UTC)

I don't think it is. - RoyBoy ⁸⁰⁰ 23:14, 20 November 2007 (UTC)

Perhaps we should review what the terms Correlation and causal relationship mean? Correlation is the co-incidence of two phenomena, but does not mean that one factor influences another (remove the hyphen from co-incidence and that's all a correlation is). A causal relationship means that one factor does influence the other (a la "adversely influence a woman's subsequent risk of breast cancer"). But that's ok - don't take my word for it:

• Lanfranchi: "and only about 5 – 10% of women with breast cancer develop this cancer because they had an abortion" [emp. mine] http://www.jpands.org/vol8no2/malec.pdf

• Brind: "Scientists know that abortion causes breast cancer but are afraid to say so publicly in today's hostile political climate." http://archive.newsmax.com/archives/articles/2002/1/4/172317.shtml

• Malec: "It's time for scientists to admit publicly what they already acknowledge privately among themselves - that abortion raises breast cancer risk". http://www.cnsnews.com/ViewCulture.asp?Page=/Culture/archive/200710/CUL20071005a.html

Clearly, pro-life ABCers think that abortion causes cancer, by increasing risks (causal relationship). A correlative relationship in no way argues for a cause and effect - that abortion is the cause of increased risk of cancer. And since 2003, no reputable org recognizes a causal relationship. They don't even recognize a correlative relationship. Phyesalis (talk) 00:25, 21 November 2007 (UTC)

That certainly satisfies me. I simply could not recall such language being used, and I'm disappointed it apparently was. Thanks for the leg work Phyesalis. I hope to reciprocate soon with Russo and Daling sources. - RoyBoy ⁸⁰⁰ 00:58, 21 November 2007 (UTC)

Thanks! Phyesalis (talk) 06:34, 21 November 2007 (UTC)

Restructure

I'm thinking we should move the NCI workshop down in the article. It makes more sense if it's after all the other studies. I also think we should allow the Beral to stand apart as its authors categorized it as a "collaborative reanalysis" - it is inappropriate to allow Brind's opinion to dictate its categorization. I also think we should have a paragraph introducing Brind, Lanfranchi and Malec as well as disclosing and contextualizing their relationships to BCPI and ABC and pro-life advocacy. We could rename "NCI" to "Proponents" take out the first paragraph move it down to "Politicization" and expand on BLM. They are notable proponents and deserve to have a little context given to their work. Thoughts? Phyesalis (talk) 06:34, 21 November 2007 (UTC)

That's how I had it initially; as the NCI is not a study my feeling always has been that it should go after the real science. Others disagreed, and as part turning the structure upside-down I put the NCI before the studies given the prominence given to it by journalists.

I agree (and have seriously considered) moving the NCI section into politicization (under its down sub-section), but I find it hard to make such a bold move without a completely reliable source (someone other than ABC advocate) to back up the political context/dynamic of the workshop. But that could just be me being paranoid of how it would look to other editors.

I wouldn't support separating the NCI paragraphs; as ABC proponents are certainly not the only ones that require contextualizing. - RoyBoy ⁸⁰⁰ 23:26, 21 November 2007 (UTC)

The only documented "political" aspect of the NCI workshop is that it was a response to Bush Administration efforts to play up a supposed ABC link. No reputable source, other than perhaps Brind if you consider him as such, has alleged that the NCI's conclusions were politicized - and Brind gets more than enough airtime in this article. As to "real science", that's a bit judgemental. After all, WP:V and WP:OR actually suggest that we should base the article on expert interpretations of the primary data (such as the NCI panel's conclusions), rather than relying on our own personal parsing of the primary data. MastCell ^Talk 06:16, 22 November 2007 (UTC)

Agree that the NCI's political aspect is only in the context of legal/political interference, but I think it's critical. NCI wouldn't have had to call a consensus workshop if pro-lifers hadn't started lawsuits and the pro-life administration hadn't started messing with their website. We don't have to separate the two paragraphs, but take the pro-life info about Brind out (I think its just part of one sentence) and bring it back up in a paragraph about BLM and the orgs - ABC was partially founded in conjunction with "Conservative Christian Women" (can't remember the name right now - ugh - been baking pies for too long) pro-life, anti-gay, etc.. I think the chronological order works. Additionally, I don't think we should be peppering every study with Brind's commentary (undue weight) - it can be summarized (and would flow better) if we put it all together with his meta-analysis. Here's a gov't article we can use to balance out the ABC advocates take on the workshop.

• http://oversight.house.gov/features/politics_and_science/example_breast_cancer.htm
• http://www.usatoday.com/news/health/2003-02-26-cancer-usat_x.htm
• http://www.publiceye.org/magazine/v20n2/v20n2.pdf (B.A.D. Science p.13)

So, MastCell, is that a yes for moving it down? Phyesalis (talk) 10:39, 22 November 2007 (UTC)

There was nothing fundamentally scientific produced at the workshop, by their own summary they reviewed the scientific evidence; and then reasserted the forming scientific consensus on the ABC issue. It created an editorial, by many experts, but an editorial nonetheless on ABC research up until that point. There was nothing published and signed providing credit/liability to their findings.

And a big "hellz no" to extricating his Brind's commentary and including it with the only science he has published on ABC. Ick! They are simply not interrelated; you can remove undue weight by finding responses to Brind... otherwise I'm very resistant to removing/jumbling together the minority voice. It could work at we create a Brind "Criticism" section at the end of the Science section, but that may be giving even more weight to Brind given he only has a few points (per study) worth consideration. Also, to further counter the notion of undue weight, in some criticisms (as I discovered recently, with Michels on Melbye misclassification) other scientists make those some of those points as well. - RoyBoy ⁸⁰⁰ 00:07, 23 November 2007 (UTC)

We may have fundamentally different views of "science". It's not just generating primary data. A critical aspect of medical science is reviewing available primary data and synthesizing it into usable, applicable, practical conclusions. That's what the NCI workshop did, and it's just as "scientific" as the NIH guidelines on cholesterol, or HIV/AIDS, or anything else. I do think Brind is given way too much WP:WEIGHT here... I'll have to dig it up, but in one of Brind's articles he essentially says flat out that every reputable or authoritative medical and scientific body has dismissed the ABC link (of course, he used the word "denied") - but the point is that there's really no question that his is a tiny-minority viewpoint scientifically speaking. Certainly notable enough to cover, but the layout of the article makes it sound like Brind and the rest of the scientific world are on an equal footing here and ought to go point-counterpoint - which is where the WP:WEIGHT issue comes in. MastCell ^Talk 03:08, 23 November 2007 (UTC)

My problem isn't with criticism of the studies themselves, independant sources for the crit are fine. I just strongly object to using Brind et al's characterization of other scientists' opinions or hiding their opinions in under the anonymous guise of "critics"(I'm still not sold on sources like "Frezza" and "etters" as valid despite lowering the bar in view of fringe theory status - if any personal page or anonymous data dump is valid, I could just go make my own home page and post whatever I wanted to support an argument). It's fine to use unreliable sources to expand his theory, but not to use them as a source of general or mainstream scientific criticism. Furthermore, they are interrelated as his criticisms are the basis for his study. As the article stands it gives his position too much weight. We get Brind's commentary on the Lindefors-Harris study in the sub-section on Beral, which is two sub-sections before the LH study is even introduced! I mean, really? It's a survey of studies, not "a survey of studies through the eyes of Dr. Brind". Is there a more compelling reason other than the assertion that his crit and his study aren't interrelated? Or, why do you think they're not related?

Ah - edit conflict - MC, we're saying the same thing. The article needs to be restructured. Additionally, the deeper I get into this article, the more I think it needs delisting or reassessment until its multiple issues can be addressed (The primary reason being that the deployment of certain ABC sources in this article borders on academic dishonesty. Not "good".) Phyesalis (talk) 04:12, 23 November 2007 (UTC)

Fine, do what you will; just don't lump it in with his meta-analysis. Even though he makes a lot of these points in that meta-analysis, it would be confusing to shove them all in there.

I think I fundamentally object to this because there isn't anything to say Brind's points are marginal. While Brind is, that doesn't necessitate all his criticisms as marginal (even if you dig some up from here and there). When Michaels pointed out the same thing Brind did about Melbye, that didn't suddenly make Brind's point any more or less valid or verifiable. - RoyBoy ⁸⁰⁰ 05:00, 23 November 2007 (UTC)

Could you please restructure your Issues list, the references do not match up. Please list issues with a section reference. Addressed issues should be Strike-through text so that we know they have been addressed. I was trying to address another easy issue, but couldn't figure out precisely what it was referring to anymore. - RoyBoy ⁸⁰⁰ 05:10, 23 November 2007 (UTC)

No problem. I'm afraid I had to cut my teeth on this article (my first major "critique" of a page, if you will) and the way I wrote it out isn't the most useful. Sorry about that. I started striking out text. Will move on to recategorizing. As far as the restructure, I'm thinking some of the info (that which fits chronologically or makes sense) can stay where it is, but by the same logic, some will fit and make more sense in with his study. Yes? Phyesalis (talk) 07:00, 23 November 2007 (UTC)

Simple and logical, sounds perfect. Almost too perfect! :"D RoyBoy ⁸⁰⁰ 23:49, 24 November 2007 (UTC)

Refrain from adding "experts rejected" every chance you get. It gives the inaccurate impression all experts reject the ABC hypothesis. This is not the case; and there is some evidence of an ABC correlation, indeed there is more than some... stop framing it otherwise. (ie: This alteration, which suggested that there was some evidence when experts rejected the theory, prompted an editorial in the New York Times) Again I repeat, there is more than some evidence. Implying otherwise is POV. Sub-sectioning NCI, it is an important topic... and is mentioned in the lead. - RoyBoy ⁸⁰⁰ 02:57, 27 November 2007 (UTC)

Also Brind commentary on the NCI workshop is out of chronology, place and I think provide more / too much weight being that high up in the article. It should simply go with the NCI, as Brind has more of substance to say than the editorial: "This issue has been resolved scientifically . . . . This is essentially a political debate." That epidemiologist confuses scientific paradigms/consensus with science; how depressing... and certainly appropriate for the section it is now in. - RoyBoy ⁸⁰⁰ 03:17, 27 November 2007 (UTC)

First, facts are NPOV. It is unfortunate (for some, I guess) that this has been so overwhelmingly dismissed by every major reputable org and epidemiologist. Second, what part of the difference between correlative and causal do you not understand? I apologize in advance, but I feel common sense requires me to be somewhat blunt. I have already pointed out repeatedly the difference between the two, provided documented support for the fact that ABC is a causal hypothesis, not correlative and documented proof of the fact this has been rejected. Evidence of correlation is not a correlation, because none of the evidence of correlation was statistically significant (fixed that in the stat bit in "Cohorts"). Therefore evidence of correlation is not support of ABC. ABC had to first establish a correlative relationship, then establish the mechanism of causality and then prove that. While it alleges a mechansim of causality it has not proved a correlative relationship and therefore cannot ever hope to prove a causal relationship. What part of this do you not get? It seems like you are really pushing OR or POV on this, but I hesitate because you don't seem to have a clear understanding of the scientific material presented. Maybe you really just don't understand it. All experts really do reject a causal relationship between abortion and breast cancer. Only pro-life fringies support the hypothesis and since none of them are epidemiologists and all of the are working outside their field (Carroll isn't even a scientist - he's a freaking actuary for christ's sake!) they do not qualify as experts. Experts are NCI, the head of the epidemiology, umm...just your average epidemiologist is more of an "expert" than these pseudoscientists. Show me one REAL expert (someone who is a) an epidemiologist b) well respected in his or her field, c) has published repeatedly on the subject of cancer epidemiology and d) holds a prominent position in a respected organization) who has found there's a causal relationship and has published to that effect since 2003. Just one Because there are none. I will not stop framing it as "rejected by experts" until you can provide one single shred of evidence to support your claim to the contrary. Without proof, and a mounting pile of evidence to the contrary, your opinion is baseless OR. My assertion of the facts has plenty of documentation from top tier sources.

Perhaps we should move this discussion to a user talk page? Phyesalis (talk) 03:40, 27 November 2007 (UTC)

You are a partisan editor and have a partisan understanding of the science and what the consensus actually is. Daling is an expert in the field, she and her research does not reject the hypothesis. Indeed no one rejects the hypothesis at all. The closest I've ever seen, is the "wobbly" comment, even there he does not state "the ABC mechanism is wrong, its a misinterpretation of etc etc"; he doesn't say that because it isn't accurate. They do reject an ABC association (that's why I'm not reverting you en mass as yet), but no one comes out and says the mechanism (the hypothesis itself) is wrong; rather that the epidemiology does not support it. It is you that is confusing political commentary (calling ABC pseudoscience) with actual science and what real scientists are actually saying. Do you honestly think you have a handle on this subject? Just because you are improving the article (thank you) doesn't mean your POV is perfectly attuned to reality and the nuanced difference between rejecting an ABC link, and rejecting the ABC hypothesis. Rejecting the link has happened, rejecting the hypothesis has not.

I'm being patient with you. Confusing Wikipolicy and 1990 vs 1999 references are not the only errs in judgment you have had on this article. When your editing is brisk, my response will be brisk if you overstep. Try to not be surprised. - RoyBoy ⁸⁰⁰ 04:07, 27 November 2007 (UTC)

And you saying keep focusing on "causal", even a correlative relationship (which has been shown in studies), is in line with the ABC hypothesis. As explained in the mechanism section. I'm unsure how you've come to believe you can define the ABC issue as an all or nothing (causal or nothing) proposition. - RoyBoy ⁸⁰⁰ 04:10, 27 November 2007 (UTC)

This is certainly a situation I need to research into further; as the lead stipulates the hypothesis as "causal" only (whereas Mechanism situates more accurately as a indirect correlative factor), and regardless of what ABC proponents say the ABC hypothesis isn't merely defined by them. It also has its origins in very cautious and conservative interpretation of Russo and Russo (and others previous to them), and speaking off the cuff here, I do not think they deemed it to be causative. They've noted a weak correlation in the past and for Russo "essentially" the same risk as virgin rats. I see what happened here, you've taken the pro-life definition of ABC as the only definition. That simply isn't the only way to define the ABC issue.

This is something that should be directly addressed either in the Lead or the Mechanism section. You've inadvertently revealed a key misconception about the ABC hypothesis that hasn't been address in the article. Now this is what I call progress! - RoyBoy ⁸⁰⁰ 05:06, 27 November 2007 (UTC)

This is ridiculous. RoyBoy, no of course they didn't deem it causative, because they were not proponents of the ABC hypothesis. None of the causative theory really shows up until Brind and the pro-lifers. Prior to that it was all just studying to see if there was any reliable evidence for a correlative relationship, for which none has been found. Again you misunderstand how causative and correlative relate, and how "weak evidence of a correlation" does not mean "a weak correlation". And while your magnanimousness is something to behold, try not to be surprised when I tell you that my patience is growing thin. My alleged "partisan" nature has nothing to do with the facts, the quality of my sources or:

Some aspects, especially those listed below are indeed ridiculous (sometimes on my side, but on your side as well). Daling is reliable, stipulating "none has been found" isn't accurate. While I appreciate where your weak evidence is not a weak correlation criticism is coming from; why should one dismiss a weak correlation in the face of conflicting evidence. Don't try to tell me there isn't any, because that is ridiculous. I nor Wikipedia is forced to see the NCI as definitive; and you would be foolish to do so.

• your inability to understand basic frames of references like correlative as opposed to causal, or the relationship between relative risk and confidence interval, leading to OR interpretations of studies.

That's silly. While the prose of the article might have conflated them; that is my poor writing not ignorance. I understand ABC proponents posit a causal link, I also understand reality might be different than they, or you, prefer.

• your OR ability to quote a study's conclusions based on one table that mostly likely comes from a pro-life website (you know, the Russo study YOU can't find in its entirety online?)

As stated, the Nature archive doesn't go back to 1987. I have made it a point to put the pages I do have in my bag; you can rest assured your assumption of bad faith won't reflect well.

• your inability to do even the most basic fact checking (Russo, Jasen, confidence intervals, rat breasts)

Kettle black. You've quite the gall to list that given your recent spat of false positives and semi-original research. (I say semi, because it was likely just sloppy writing on you part, specifically not prefacing pregnancy benefits with "full-term". Removing "egregious" from its context is was also poor form)

• your ineptitude or gross malfeasence with the first Russo study in which you directly misquoted the authors conclusions even though they were clear as day,(you know where you quoted a paragraph on benign lesions in support of your BS assertion that the Russo study found abortion to cause cancer in rats and tried to refute me with OR). Really, I just want to know, how did you read that whole study and come to the conclusion that Russo and Russo found an increased risk in cancer? I mean, it says they didn't right in the summary!

Well because unlike most people I did read the whole study, not just to the summary; and I came to understand there are higher carcinoma levels for Pregnancy Interrupted rats. I'm still curious how you've come to think Russo and Russo refuted ABC when their studies do nothing of the sort. But why should I tell you this, apparently all you need to know is in the summary; unless of course the summary disagrees with your OR of Howe's results.

• your extremely "partisan" dismissal of peer-reviewed scholars who do not support ABC like Dr. Mor, when you removed sourced material from an expert that rats did not have breasts, and Jasen who showed how the Russo study is often misquoted in support of the pro-life ABC hypothesis (as it was being used here) which all suggests that your research has come from unreliable pro-life websites.

That was my mistake, as I did not remember that being in the source. (which I found, and I added the Rats are not Humans segment as well by the way, long time ago)

• your acknowledged OR to a support spurious pro-life quote

Spurious? Says who, you? Since Daling acknowledge the latter sentence, I'd hardly consider you to be in any position to consider it spurious. You have a healthy skepticism of pro-life sources, don't let it get the best of you.

• your inability to understand that the causal relationship is rejected because it hasn't been able to establish a correlation, which is the first of three important steps in establishing causality - if there is no correlation there can be no causal relationship, and that while there is old and marginal evidence that suggests a possibility of a correlation, this does not mean that there is a correlation, particularly when new studies with better info and methods have definitely proved that there is no correlation. You have no idea what you are talking about. I have sources to prove what I'm talking about.

You're being selective and transparently partisan here. If you continue to maintain there is "old and marginal evidence that suggests a possibility of a correlation" I won't accept you have the ability to maintain a neutral perspective. Howe isn't marginal nor old, Melbye's and Michael's positive sub-groups aren't either. They are insignificant, but these are hardly marginal results from no-name studies.

Further, response bias is an unsupported assertion (you simply cannot conveniently ignore all peer reviewed interview studies en mass); and the studies that have attempted to show response bias statistically significant are laughable... and how they were received uncritically was disturbing and frustrating to me.

all of which you have done in order to support of the rejected pro-life theory of ABC. And you suggest that I am partisan? Let's be honest, since you have cast more than one aspersion on my "nature", if anyone is partisan, it's you. You have used unreliable sources and acknowledged OR, you have written that you intended to continue to engage in OR to refute the proper quoting of Russo et al.'s conclusions, and you have done it all to support the ABC hypothesis. Acknowledging the well-documented facts is not POV. Abusing sources and relying on unreliable sources/OR to contravert established facts is POV.

I go the extra mile to try and find the truth of an important quote from a researcher who seemed quite fed up with the likes of us misappropriating her science, and it gets used against me. I removed it as a show of good faith, and you continue to HARP on a quote/point you may not even be correct about. Sheesh, fishing on a low road me thinks. And I wasn't quoting their conclusions, I was attempting to communicate their actual results. I continue to be befuddled at how "similar or even higher incidence of benign lesions" and carcinomas" is incorrect or OR. It doesn't match their conclusion, okay, but it does come from their study directly... and admittedly indirectly from their data.

I'm stepping back. Until you can provide reliable sources to prove your assertions, there isn't much to talk about. You can use unreliable sources to explain fringe theories, but you can't use them to contravert documented consensus. Until you find something from a reliable source after 2003 that explicitly states that scientific consensus has changed and is once again considering a causal relationshiip, you are just asserting OR. This seems to be a habit of yours. I have mentioned this several times before, all politely. If you can't contain your opinion and exuberance for the issue (or at the very least couple it with reliable sources), additional steps may have to be taken. Phyesalis (talk) 05:16, 27 November 2007 (UTC)

I've allowed you more than enough running room to make mistakes of your own; and you've made some small ones; as have I. But on the point of "causal", possible correlative relationships data cannot be ignored and minimized on your, or NCIs, say so. And don't try and paint me into a causal corner, heh... causal corner... because I'm not that narrowly focused nor am I an epidemiological neophyte. They (NCI) do not have a dominion on the science, in fact you cannot even be confident of what science and data they choose to consider at the workshop! (unless you can, in that case, gimme the source!!! cause I would like to read it)

But I gotta hand it to ya, stepping back is a veteran wiki-move; I plan to reward you with sources. (chart papers are literally in the bag) Would you prefer I e-mail them or post them here, what format do you prefer, TIF or PDF? PDF is likely best. - RoyBoy ⁸⁰⁰ 23:48, 27 November 2007 (UTC)

I'll take full articles in PDF, incomplete info is useless, as I question your bias. E-mail, thank you. I'm no veteran (I've spent a good portion of my time since I started on WP out of the country), I just read the policies and guides before I started editing. I may have made some small mistakes, but OR (for "important" quote or not) and totally contradicting the Russo findings are not by any means small mistakes. Let's not confuse typographical errors with ignoring an author's conclusion and engaging in OR analysis to make the quote state the opposite. Ok? Only people with agendas or inept neophytes make that mistake. So let's just cut the crap. You question my ability? You can't even read a study summary, let alone quote it without letting your bias show. All my inclusions are from reputable sources. It's funny, you'll accept anything that comes from pro-life newsletters as gospel truth yet you question the NCI, the American College of Obstetricians and Gynecologists, the Royal College of Obstetricians and Gynecologists, Netherlands Cancer Institute, New England Journal of Medicine, AND the head of Epidemiology for the American Cancer Society? And I'm biased? Maybe you should step back and rethink things. Phyesalis (talk) 03:02, 28 November 2007 (UTC)

Opposite? Hmmmmm... not really. I doubt I have to rethink that part of it, as it was a well considered opinion based on observational evidence of you having a habit of unconsciously situating/contextualizing opinions (of others based on good sources) as facts. Anyway, as I did previously mention, I am in possession and have PDF'd two pages of the study. I will have to make time on the weekend to acquire it in full; I'll send it to Mastcell for kicks. - RoyBoy ⁸⁰⁰ 22:12, 28 November 2007 (UTC)

Not open on the weekends, so may have to head over to the university... but that's a hassle! And I have to get the copy card with the monies (unless there is a way to PDF it in the library these days... hmmm... doubtful, on the up side I could give the LA Daily News a college try at the same time) - RoyBoy ⁸⁰⁰ 23:03, 28 November 2007 (UTC)

OK, let's all step back for a minute. It's best to un-personalize these things as much as possible, otherwise bad blood ensues. There is a significant content issue here, though. RoyBoy quite correctly points out that studies cannot be minimized on our say so. However, the NCI is different - it's an expert organization of cancer specialists, and their conclusions about the relative worth of the various primary studies should carry quite a bit of weight on Wikipedia. Again, we're just trying to represent opinions on the matter in proportion to their representation among experts. The members of the NCI panel are experts. We may have our own opinions about their political bias or lack thereof, but their conclusions about the relative weight of the various studies are highly relevant for Wikipedia's purposes. We can speculate about what data they included, and why they weighted things as they did - in fact, it's an important part of being scientifically literate. But our own conclusions on the NCI panel findings, however well-informed, don't really have a place on Wikipedia, nor should they eclipse or marginalize the NCI's actual findings. Of course the NCI does not have a dominion over science; there are any number of other highly respected scientific/medical organizations, yet they all seem to agree that an ABC link is bogus. That needs to be a bit more clear here, as the current layout of the article gives the impression of a more robust scientific debate than the sources support. MastCell ^Talk 05:48, 28 November 2007 (UTC)

True. - RoyBoy ⁸⁰⁰ 22:12, 28 November 2007 (UTC) Wait a second here, historically I wouldn't agree with the concept of weighting the NCI. We do not know the experts individual opinions and insights into the ABC issue, the NCI and the evidence. I think we should give them as much weight as we can verify; and who says the experts had significant input on the weight of the evidence? There seems to be an small inference they went into a democratic intellectual exchange at the workshop, based on what I've read from Brind, and what I've seen/know from documentaries on governmental oversight of conferences... that is not the case. Speakers, subjects and information/data is set by those organizing the conference, and if you'd like to dissent or bring in new evidence; you need to be prepared for an uphill struggle. And I'm not thinking of ABC when I say that; but I can imagine politically hot topics are even more unattractive. - RoyBoy ⁸⁰⁰ 23:03, 28 November 2007 (UTC)

The NCI is the pre-eminent cancer research organization in the US (and one of the pre-eminent in the world), and it was chaired by a Bush appointee (and plenty of reliable sources attest that the Bush Adminstration is not shy about putting its thumbs on the scales of science for political reasons). If anything, one would expect political interference to lead to the opposite conclusion, based on reliable sources and history. Has any reliable, independent source (other than Brind, whose opinions are amply documented here) alleged that the NCI workshop's conclusions are suspect? If it's just our own gut feelings about the NCI, then while those may or may not be valid, they don't really have a bearing here. MastCell ^Talk 00:31, 29 November 2007 (UTC)

No independent source has questioned the NCI conclusions. However, you are defining the issue in a troubling and backward way. The question, from a scientific perspective should be, has any independent organization peer-reviewed its results and methods? While I understand you are emphasizing their "pre-eminent" status to validate your point about weight; I couldn't care less who they are and I regard my relative ignorance about them as an asset. I want to know what they considered, how they considered it and who was at the podium speaking for "the workshop". Thusfar I've come up with Jack (excluding Brind); and I gave up trying a long time ago. If you and Phyesalis have more than your gut feelings, then educate me. Otherwise my gut feeling is based on more than yours; since while Brind is partisan he is no fool, he is an expert on the subject and on the proposed mechanism no one has yet thrown out. (despite assertions to the contrary)

On the issue of the Bush appointee, I was quite harsh to Pro-Lick for rendering a similar facile political analysis of a very large and well establish progressive scientific organization. It is a red-herring and I sincerely hope you reconsider it. - RoyBoy ⁸⁰⁰ 02:43, 29 November 2007 (UTC)

Most importantly I should note, that since so much is unknown (to me) about the NCI workshop; I've done my level best to infer what I can. The NCI saying there is a "gap" for pre-term delivery and breast cancer risk, when there was a Melbye study available. What does that say? Likely they considered the Melbye study too insignificant to mention, but it also indicates they are willing to minimize/ignore evidence which directly contradicts their well-established findings. I also wanted to mention I removed a Brind line from the NCI criticism, as I found it to be an obvious and vague criticism; which detracted from his relevant and verifiable points and provided better weight (as requested/pointed out by Phyesalis and MastCell) to the section and article. I am trying guys. - RoyBoy ⁸⁰⁰ 03:05, 29 November 2007 (UTC)

RoyBoy, the fact that you "couldn't care less" about actual experts and continue to rely on the opinion of a disreputable pro-life activist is troubling to me. Sufficed to say, other experts (like the ACS, RCOG, ACOG and Netherlands Cancer Institute) have found it acceptable and that is all that's required for WP. Your opinion is irrelevant. A number of studies have come out post-NCI that have repeatedly shown no correlation. I realize this is your baby, but just because you started a page does not mean that your view is valid. You don't WP:OWN this article. I suggest you stop pushing POV about NCI and "experts". It is disruptive and time consuming. Both MastCell and I have been over this with you several times. The theory has been rejected, experts find no causal relationship because they find no correlation. I think a lot of this stems from the fact that you fail to understand that relative risks of 2.0 or less are statistically insignificant. This is where starting a page primarily sourced by "etters" the pro-life homepage can get you in trouble (which BTW should be stricken as a source - there is no excuse for such a poor quality source, fringe theory or no). If you would like to dispute the NCI findings, you need to find a better source than Brind and co. Your position is the uphill battle. We do not have to prove anything, you do. I appreciate your continued efforts. Thank you. Phyesalis (talk) 04:19, 29 November 2007 (UTC)

I should clarify I don't care about credentials when no science is presented, I've found credentials and peer-review to be wrong more times than I can count, so I've learned not to rely on them as much as others. Remember this letter I posted for the Carroll issue. The author is the Head of Surgical Oncology, and it is entitled "Women have right to know correct data". But then no data is actually provided. This isn't an oversight, but rather a habitual reliance of argument from authority I see frequently in lay publications; in the scientific literature this does not occur, and there is a much better (honest and open) dialogue on the evidence at hand.

There are no declarations of proving ABC flawed... rather that it is unsupported. The NCI did cross that line with a robust "well-established", and rejected a causal link. Fine, but their findings are not in keeping with the scientific dialogue; nor do they provide the data to back themselves up. It, in my current opinion, is an elaborate version of a letter to the editor. I would also point out, one should take care to situate "rejected" as an NCI determination; not a scientific fact.

Brind's dis-reputation is partially a result of slander, pseudoscience and subjective as a result. Yet... he overstates the evidence, yes, he gets things wrong, yes, he isn't neutral, yes. Please tell me how your experts are exempt of these flaws, because I assure you they are not. Is their position better supported by the evidence? Absolutely. If you agree that the evidence should be King, then the NCI workshop is a Prince at best and a court jester at worst on the ABC issue. Man, I earned my supper with that line! (eats pork, watches BBC News, orders Canon SD850 IS with a sweet online deal, returns to Wikipedia)

The evidence indicates a weak correlation sub-groups (no correlation is a tenuous position, reminder, the conclusion in the abstract isn't the only thing Wikipedia should be interested in; further Howe's and Melbye's abstracts indicate correlations); it should be clearly communicated a causal link has been rejected by the NCI and current studies, and doesn't even impact the ABC hypothesis (as I define/understand it), though causality is certainly key to pro-life's ABC link campaign. It should also be made clear the hypothesis itself remains plausible in specific circumstances; as indicated by the science. The NCI, I reiterate, does not change that.

Uphill battle? I never would have guessed. :"D RoyBoy ⁸⁰⁰ 00:15, 30 November 2007 (UTC)

I think the bottom line is this: your interpretation of the primary data, no matter how well-founded, does not trump the interpretations of the National Cancer Institute, American Cancer Society, and every other major medical and scientific organization that has commented on the issue. These aren't "my" experts, as you suggest above. They're the experts. MastCell ^Talk 04:47, 30 November 2007 (UTC)

Absolutely, and the experts have data showing correlations; that I think is part of the bottom line. But there is something else going on here, it isn't just the interpretation and weighing of evidence that is getting me so defensive and argumentative; but there is an effort to say ABC is bad science. That is not reflected even in the NCI. They look at the evidence and say its unsupported, but they do not stipulate the ABC mechanism is flawed, a misinterpretation, or pseudoscience. Phyesalis is confusing the pseudoscience pro-lifer's put out there (and get golden boobs for), with the real science behind the ABC hypothesis. - RoyBoy ⁸⁰⁰ 06:03, 2 December 2007 (UTC)

NCI website timeline

I was trying to find a link to support the "NCI updated website" as a result of pro-life congressman poking around in 1998 (Tom Coburn) and in 2000 (Tom Bliley), and I found a really bad source.[4]

Anyway, I tried to get better sources and got a differing narrative on the NCI website editing; than what we have in the article. Sciencemag.org and a partisan Washington Monthly specify that specific congressmen were involved in pressuring the NCI to change its website in 2002; rather than the Bush administration.

So even though the source for 1998 to 2000 is REALLY bad, I think it has enough truth... and I'm trying to figure out how to confirm it... and if it turns out to be semi-veriafiable, how that ties into (if at all) with a longer and more protracted NCI timeline; which also Ties into the North Dakota lawsuit and its use of NCI material.

Essentially it might be inaccurate to say the Bush administration was responsible, though maybe it was for specific edits in November 2002. I'm suddenly unsure. - RoyBoy ⁸⁰⁰ 04:32, 27 November 2007 (UTC)

Intro

Okay this is what I took out of the intro:

A large epidemiological study by Mads Melbye et al. in 1997, with data from two national registries in Denmark, reported the correlation to be negligible to non-existent after statistical adjustment. The National Cancer Institute conducted an official workshop with over 100 experts on the issue in February 2003, which concluded with its highest strength rating for the selected evidence that "induced abortion is not associated with an increase in breast cancer risk." In 2004, Beral et al. published a collaborative reanalysis of 53 epidemiological studies and concluded that abortion does "not increase a woman's risk of developing breast cancer." However, these studies may suffer from misclassification and have incomplete histories of womens reproductive years. Dr. Brind argues some interview-based studies indicate a statistically significant correlation. Debate remains as to the reliability of these retrospective studies because of possible response bias.

This material could be worked back in (the citations have to be put back in) - RB - you around? Got any suggestions? The intro now covers an overview of all the material covered in the article. Maybe we could summarize the Brind material into one sentence, but the intro covers the theory, some historical points in the development, address the legal issues and helps contextualize the importance of the 2003 NCI workshop. What do you think? Phyesalis (talk) 06:15, 21 November 2007 (UTC)

I am now for a little while. I'm simply unsure what you are trying to accomplish. Was the middle paragraph not flowing well itself, or in relation to the other two paragraphs; or are there redundancies you want to minimize... or are you looking to update it as well? - RoyBoy ⁸⁰⁰ 23:30, 21 November 2007 (UTC)

The Brind sentence should be merged, but last night I was too tired to completely trust my grammar skills. - RoyBoy ⁸⁰⁰ 23:40, 21 November 2007 (UTC)

I do not understand why this remains outside of the article. I will try and make it shorter, and then reinsert it into the article. An article of this size necessitates 3 solid paragraphs in the lead. This paragraph is meant to summarize the science; the others Intro the subject, and describe the current ABC situation. - RoyBoy ⁸⁰⁰ 18:04, 25 November 2007 (UTC)

A lot of changes, most of them fine. However "(none of whom are epidemiologists)", brackets are bad style and irrelevant. Does one need to be an epidemiologist to render a meaningful opinion on the ABC issue? Of course not, its an argument from authority... you do that plenty already. Questioning credentials is fine in the article (it has little relevance anyway if someone, such as Brind, does have relevant expertise), but it stretches my imagination to see how its notable enough for the lead. I found it borderline to even list Malec and Lanfranchi in the lead; let alone start a biography on them. - RoyBoy ⁸⁰⁰ 02:09, 27 November 2007 (UTC)

I am also reverting informed consent, which I believe to be my best addition to the lead in quite some time. I summarizes an entire paragraph regard state laws in half a sentence; it also clarifies, regardless of how many times rejected and refuted is added to the lead... that even the hint of a risk, might possibly be libelous depending on the jurisprudence of informed consent in a specific state. (as, to my understanding, they vary slightly) - RoyBoy ⁸⁰⁰ 02:16, 27 November 2007 (UTC)

First you rewrite the intro to Mechanism which implies pregnancy is automatically a "benefit", which isn't the case as Russo and Russo have indeed refuted for rats. Without evidence to the contrary there is no reason to assume human pregnancy is immediately beneficial to breast cancer risk. Then you rework it to say ABC misinterprets (implying its flawed and/or pseudoscience) a beneficial mechanism, which you have asserted for early pregnancy! Tread more lightly. - RoyBoy ⁸⁰⁰ 02:31, 27 November 2007 (UTC)

I don't even know what to say. I reworked a non-controversial opinion: pregnancy confers a protective benefit. This is well-documented. ABC does misinterpet this as having a correlative negative effect: namely that abortion causes an increase in one's risk of cancer -this is sourced by the Jasen paper. Really RoyBoy, maybe you should step back for awhile. ABC is flawed science and we have provided multiple citations (Mooney, Yeoman, Jasen, the article posted here above on B.A.D. Science, the fact that it won the Golden Boob award from the National Breast Cancer Coalition Fund for bad science) that assert this. Please consider taking a break. You do not have a clear or unbiased view of this subject. Phyesalis (talk) 00:59, 30 November 2007 (UTC)

Pregnancy does not confer a protective benefit.

Full-term pregnancy confers a protective benefit.

Does clarify what I am trying to communicate? - RoyBoy ⁸⁰⁰ 03:50, 30 November 2007 (UTC)

Oh yes! Sorry, I thought that I had acknowledged that before. It was a typographical error in the original edit. You are correct. I did intend "full-term" pregnancy. I'm sorry. I'm certainly not objecting to you changing "pregnancy" to "full-term pregnancy". Have I suggested such? But that doesn't address my main point. You continue to assert that ABC isn't bad science. It is. ABC does interpret the benefits of "full-term pregnancy" to have a correlative negative effect in the form of abortion increasing the risks. The problem, as the Russo studies continue to prove, is that while PI rats have higher risks than parous rats, they still have the same risks as the baseline virgin rats. Phyesalis (talk) 04:03, 30 November 2007 (UTC)

I don't understand, an aborted pregnancy is not a full-term pregnancy. I have reverted it again, misinterprets what? Who says it misinterprets it, where are you getting this from? - RoyBoy ⁸⁰⁰ 22:33, 2 December 2007 (UTC)

Well I fully acknowledge my previous interpretation of Russo has been misleading, especially assuming that the P.I. rats are a spectrum of ages, at the higher carcinoma level (relative to O.V. rats) is mostly the result of younger ages in the P.I. sub-group. Despite being marginal results; are they marginal higher or lower, and are they consistent?

Now, ABC is plausible (good and bad are editorial words I try to avoid) science and hence continues to be a hypothesis because:

• Hormones play a key and complex role in breast cancer.

• Hormones play a key and complex role in pregnancy, especially early on.

I refuse to ignore the real possibility that growth initiated by pregnancy has zero effect on breast cancer risk if interrupted. Unless, through mechanisms such as hCG, it can be demonstrated that the growth factors (potentially cancerous) are actually mitigated and managed by differentiation factors (potentially anti-cancerous). You used the word ridiculous before, I frankly find it ridiculous for anyone to look at the biology of early pregnancy and say off-hand: "Naw, it has no net effect whatsoever, its a wash." It needs to be examined and demonstrated it has a net zero effect; I'm not going to fall for the bio-ethical trap (and cop out IMO): "If there is an effect, its too small to worry about." It's very likely too small for most women, but that is not good enough from a scientific perspective, and a clever lawyer in the right jurisdiction could argue not good enough from a legal perspective. It is actually an assertion to claim ABC is bad or pseudoscience, because that has not be demonstrated; despite the NCI's interpretation of the epidemiological results. I stress, the NCI did not attack the proposed mechanism, disprove or even attempt to cast doubt on it.

To go from saying, marginal, insignificant results, cooked results – to – no effect whatsoever, is intellectually dishonest and unacceptable on a subject that can have ramifications for women decades from when they had the procedure; and who can theoretically do something about it! I hope this helps our collaboration. - RoyBoy ⁸⁰⁰ 23:03, 30 November 2007 (UTC)

How is any of this wikipedia's problem? We're just supposed to condense the current thinking with a history that shows the development. I've been reading this on and off over the last week or two and see the same attempts to right a perceived great wrong that frustrated me last year. A theory is a theory - until evidence turns it into hard science. ABC has no evidence. Early experiments indicated a small risk but have been discounted. To play down a meta-analysis as bringing no new science to the table shows a lack of understanding of the scientific process. As for the NCI and any problems with political bias - surely the fact that European studies concur with the findings rules that out? Sophia 23:19, 30 November 2007 (UTC)

Thank you! Perhaps you'd like to weigh in on the discussion of appropriate sources below. Phyesalis 02:38, 1 December 2007 (UTC)

Nor is it Wikipedia's job to assert or support a tone of "bad science" or that there is "no evidence". It is this kind of dismissive editorializing that frustrates me. Play down what meta-analysis; the NCI workshop? What do you know about that workshop that makes it rigorous enough to consider it a scientific study? What new evidence did they bring to bare on the subject? Was that new evidence properly peer-reviewed? *sigh* Just because you're frustrated (about a valid concern) doesn't mean you have a clue when you say "no evidence". BTW, European studies also have positive results in them; but if you choose to ignore them, then I guess that isn't my malfunction. - RoyBoy ⁸⁰⁰ 05:47, 2 December 2007 (UTC)

It remains an assertion to call the ABC hypothesis bad science and to frame it as such. A point I forgot to make, is that even though pro-lifer's put out a lot of bad science, which reflects really poorly on the ABC issue; that doesn't make the ABC hypothesis bad science. Oh and "discounted"? Discounted by what, hard science? Heh, hardly. But whatever, double standards abound when you would rather just believe what you want. - RoyBoy ⁸⁰⁰ 06:10, 2 December 2007 (UTC)

Look at this article [5]. Clear, precise and gives a history of the issue . Old data is of historical interest once problems have been identified with the methodology, but it is bad science to keep drumming them up when superior methods such as the cohort studies have been used. From all your writings Roy you come across as someone who is utterly convinced of the theory of causation (and I admit it sounds plausible) but this is not wikipedia's problem and to push the view that somehow this question is highly undecided is OR. People will continue to conduct studies - not because they think they will find a link but because it is good practice to try to confirm or refute findings. Recall bias is an established problem which is why studies of the Danish type (one of the European ones I was referring to) are infinitely superior and why they are given such accord. If you don't like "dismissive editorializing" then you are writing in the wrong place - blogs are more suitable for personalizing issues. Sophia 09:31, 2 December 2007 (UTC)

I'm snowed in a bit, with a sore back... so back to Wikipedia. Well I don't blame you for thinking I was "utterly convinced of the theory of causation", as the article was/and remained imprecise in framing it; probably my biggest mistake on the ABC article. What I can say in my defense is when I made this edit I was actively trying to correct that mistake.

But you are ignoring the same things as last year. Recall bias is an established problem, but has it been established as statistically significant? From your link: "The likelihood that women who have breast cancer will give a more complete account of their abortions than women who do not have breast cancer is called "recall bias," and it can seriously undermine the accuracy of study results." (emphasis mine)

It can, but has it? No one has established that. Daling actively looked for it in her second study, and did not find it. There is a potential problem, not a verified problem. But, hey that's fine forget the interview based studies; cohorts have positive results and problems of their own from incomplete records (something Michaels points out as something that can undermine its accuracy) to not using the tried and true case-control methodology with the large (but incomplete) cohort dataset they have. Are they better, probably, larger dataset, yes, other things that can go wrong and undermine its accurate, certainly. There is evidence from primary cohort sources of an ABC correlation. Secondary reliable sources saying otherwise (following the NCI's conclusions) should be front and center in the article; but within the context of a secondary source and not to exclude verified caveats. Remember, every scientific study has its problems.

Okay, I'm going to try and lay this out clearly without getting sidetracked by the minutia above. Does any of your scientific source(s) say there is no evidence of an ABC correlation and/or the ABC hypothesis is false/disproven? As I read them, they all say there is no evidence of a causal link. I agree with that. Even with this strong statement, scientific sources are being conservative on their collective interpretation of the data; and they are prefacing their conclusions based on a direct causal mechanism. As I currently understand things, that's fine and accurate.

I need to assure myself that everyone here, regardless of their outlook on the on the ABC issue recognizes that is not the same thing as saying no ABC correlation and/or evidence. That's when I start getting antsy and writing long and sometimes clever posts. *crosses fingers* - RoyBoy ⁸⁰⁰ 18:18, 2 December 2007 (UTC)

RoyBoy, have you read the reliable source material in this article? The only study to posit a causation is Brind. The Weed and Kamer editorial points out that Brind's theory of causation is unwarranted and irresponsible (couched in polite but deadly serious terms).

• All of the studies examine the possibility of an association. The minority of the studies assert that there is evidence to suggest a correlation, though most of these few studies are hesitant to even say that. The majority of the studies say there is no evidence of an association, period.
• There is no evidence of causation because no one, except for Brind, accepts that there is an established correlation. Therefore, ABC, which argues for a causative relationship has been disproven. It is an obsolete medical theory. No reputable scientist buys it.
• This is the heart of what "bad science" is: preconceived notions, poor methodolgy, misinterpretation of previous studies, and failure to separate correlation and causation. We have sources that say Brind is guitly of all of these; we have sources calling his work "bad science" and "Biased Agenda Driven science;" and he was awarded a Golden Boob for his bad science!
• Lastly, the problem with recall bias is that (in theory) women in the control group do not report abortions. What this means is that if recall bias is established it will further prove there is no association between abortion and breast cancer. Proving it exists will not support an ABC link. Proving it doesn't exist won't support the theory either, as the evidence shows there is no proof. Phyesalis 18:47, 2 December 2007 (UTC)

You should tweak your understanding of recall bias; it rests on more women in the control group not reporting. Lindefors-Harris (1991) showed that women from the control and case group misreported. Indeed studies they are hesitant, I expect nothing less from scientific professionals trying to get it right/accurate. What the hell happened to the criticism of the Lindefors-Harris study? I'll paste it here for convenience and review. The ABC issue is broader than a causative association. (Side note, adding ref to CN added to Lindefors-Harris (1991) section, even though a Daling ref was available in the below section which has been strangely removed. - RoyBoy ⁸⁰⁰ 22:14, 2 December 2007 (UTC)

Dr. Daling (1994) found it "reasonable to assume that virtually no women who truly did not have an abortion would claim to have had one," and missing records could have occurred for a variety of reasons. With these eight women removed the error margin is reduced from 50% to 16% which severely limits its statistical significance. Dr. Brind believes the remaining 16% could have resulted from the Swedish fertility registry – where women were interviewed as mothers – which could have increased their tendency to underreport, given that a mother might not want to appear unfit. Subsequently the Lindefors-Harris obliquely retracted the 50% conclusion in 1998, but reasserted since the Denmark (Melbye 1997) cohort study found no ABC correlation the 30% increased risk in the Brind meta-analysis must be the accumulative result of response bias.- RoyBoy ⁸⁰⁰ 22:14, 2 December 2007 (UTC)

Whoops, I see it now. My bad, though it is out of chronology and its current placing introduces new narrative issues. I'll try to tweak it a bit. - RoyBoy ⁸⁰⁰ 22:23, 2 December 2007 (UTC)

Russo (1987) & L.A. Daily News

I've been unsuccessful in getting these sources online.

Nature only starts its Lab Investigation archive in 2000. I need the 1987 source for Russo which contains the chart in question pg 125 to be exact. (Russo J, Russo I (1987). "Biological and molecular bases of mammary carcinogenesis". Lab Invest 57 (2): 112-37. PMID 3302534.) But an interesting point, is results for the chart involve 2nd-trimester abortions for rats; I did not remember that, so an interesting twist and big mitigating factor in the interpretation of the results. I do have the chart and the preceding page... so at the very least I can PDF those and give them to you.

LA Daily News for the Daily quote, my online source (a university database) has the LA Times, but not the Daily News. In both instances I may have to hit the books to acquire the sources. Russo 1987 shouldn't be too bad as there is a cancer research center close to a friends house. LA Daily News, unsure how accessible that will be. Help would be appreciated on that. - RoyBoy ⁸⁰⁰ 01:07, 25 November 2007 (UTC)

OK, RoyBoy has sent me two pages from Russo 1987. First table doesn't even include stats on rats with pregnancy interruption. Fig. 43 shows risk rates for PI rates compared to virgin controls of different ages. The article states, in keeping with 1984, "DMBA administration after pregnancy interruption induces an incident of tumors similar to that observed in age-matched virgin rats, and remarkably higher than in parous rats." Again, as virgin rats are the baseline and both virgin and PI rats showed remarkably higher rates, I don't see how this contraverts the 1984 study. As long as the article wording remains the same (reflects consensus with 1984 conclusions - no increased risk associate with abortion when compared to virgin rats), and the partial document is not used to support statements to the contrary, I don't need the rest of the article. Any effort to contradict the cited assertion constitutes an exceptional claim and will require an extremely reliable secondary source. Sound reasonable? Phyesalis (talk) 06:27, 29 November 2007 (UTC)

But where it diverges from the 1982 study, is that the carcinoma level for these 2nd term interruptions is markedly higher. Is the TEBs the only thing of interest in that chart? From pg. 125: "All of the tumors developed by young virgin rats are carcinomas of ductal origin." I'm trying to understand how more carcinomas in the interrupted group is a neutral/non-notable result to its control group; my layman reading of it is that more carcinomas can lead to (though it doesn't have to) more tumors down the road. While I appreciate one carcinoma isn't equal to another (they can have a spectrum of stages and sensitivities), more is bad, isn't it? You're being reasonable if you can cogently explain this difference; and how it doesn't worry you. - RoyBoy ⁸⁰⁰ 22:50, 29 November 2007 (UTC)

RoyBoy what part of the bolded statement above do you not understand? The author clearly states that the rates between PI rats and virgin rates are similar, backing up the original study's conclusion. You cannot use your personal interpretation of a primary source to contradict the author's conclusion. It is a clear violation of WP:NOR. Your interpretation is not reliable or verifiable and is not notable because the authors have deemed it not notable when they stated that the risk rates were similar. Second, you don't understand the statistical significance of all these studies' findings. None of the findings are notable because all of them (except for one or two that were later corrected by methodological changes) have RRs with statistical significance values that contain 1 (the presence of values 1.0 - 1.9 means the findings are not significant). Please provide me with a reliable source that says otherwise.

The fact that you're concerned about me worrying over your interpretation is touching. Truly. But you can relax, I'm not worried and this is why: "virgin" means "hasn't had sex". So, the fact that virgin rats develop carcinomas is irrelevant because they are the baseline, that is to say the virgin rats haven't had sex, and therefore, are not the rats with interrupted pregnancies (are we agreed on this? I mean unless we're bringing virgin births into it...). The fact that virgin rats show higher rats of carcinomas just proves a) carcinogens in high doses cause cancer and b)that pregnancy confers a protective measure against this. In fact, that virgin rats also develop carcinomas at similar rates to rats who have had sex and had the pregnancies interrupted shows that abortion is not associated with the development of carcinomas. Come on, this is getting silly. Phyesalis (talk) 00:46, 30 November 2007 (UTC)

Well I can't relax quite yet, you are taking the trouble to define virgin :'p, but by saying "virgin rats" you are not bothering to parse between Young V. and Old V. Do you understand the implications of age matched P.I. rats having higher carcinoma levels than Old Virgins? - RoyBoy ⁸⁰⁰ 03:37, 30 November 2007 (UTC)

Wait a sec, what was the median age of P.I. rats. If P.I. rats ran the spectrum of age range between Y.V. and O.V., then the result is fine as P.I. reflects that age range. I was under the impression P.I. was age matched with O.V. Hmmmmm... - RoyBoy ⁸⁰⁰ 03:43, 30 November 2007 (UTC)

This isn't relevant to the page. All we really needed was Russo and Russo's conclusion that just like the earlier study, PI rats have similar risk rates to virgin rats. I think it is safe to conclude that the Russo studies do not assert a correlative link between abortion and breast cancer. We have a reliable and verifiable secondary source backing up this conclusion and we have no reliable sources to the contrary. Please let this go. Phyesalis (talk) 04:09, 30 November 2007 (UTC)

Even if it were shown that Russo rats are the same (a conclusion does not provide detailed data), I cannot let go of some results that indicate a correlation. You've stated elsewhere, some of those results were subsequently adjusted (I assume we are referring to Melbye adjusting their results at the NCI workshop). While that is all well and good, it is cooking the numbers. Is it cooking the numbers for accuracy, or because the result isn't what they expected; and they believed it needed to be corrected? While this socio-analysis gets us nowhere fast, I would reassert Michaels cannot exclude some sub-groups as possibilities. There can be no high risk groups to a non-existent/"rejected" correlation. While weight in the ABC article has and should continue to be adjusted as per Mastcell; I suspect under your hand the pendulum will swing too far and with unwarranted certainty. I might decide to relax this weekend; in order to do so I'm ignoring the article for a bit. Happy editing until then! - RoyBoy ⁸⁰⁰ 22:26, 30 November 2007 (UTC)

Ok, have fun this weekend. I just want to state that your opinion is OR. Leave it at the author's conclusion. Your analysis is not relevant. Phyesalis 02:36, 1 December 2007 (UTC)

Their findings are always relevant; and aren't always in the conclusion. - RoyBoy ⁸⁰⁰ 05:51, 2 December 2007 (UTC)

CLINMED comments

I have posted a request over at CLINMED for other users to come and lend their opinion to the issues:

• The article is on a fringe theory. The fringe theory has been dealt with in numerous reliable sources. What weight should be given to info cited in personal pages, amateur partisan websites, religous newsletters and self-published sources (see examples below)?
• To what extent should an editor intrepret tables, and to what extent should these interpretations be used to present studies contra author's stated conclusions?
• What is the statistical significance of relative risk figures with confidence intervals containing values of 1?
• Given that consensus states abortion is not associated with increased risk of breast cancer, to what exent does Carroll's prediction (Carroll is an actuary and publishes in JPandS) section claiming an increase in breast cancer rates based on abortion constitute an exceptional claim (one that would surprise the reader)? If it is, what kind of exceptional documentation does it require?

Source examples for scientific criticism:

• "etters" http://www.etters.net/cancerTP.htm#3 - this is its homepage: http://www.etters.net/
• "Frezza" http://web.archive.org/web/20041011131501/http://www.iol.ie/~hlii/beforeyouchoose.html - this is its homepage: http://web.archive.org/web/20041009151038/www.iol.ie/~hlii/mainpage.html and the article has not been located elsewhere yet.
• "Catholiccitizens.org"(state-wide) http://www.catholiccitizens.org/platform/platformview.asp?c=9446
• Self-published, Brind from his self-created BCPI: http://bcpinstitute.org/beralpaperanalysis.htm —Preceding unsigned comment added by Phyesalis (talk • contribs) 05:53, 29 November 2007 (UTC)

Recent edits and sources

Last night I switched more info around to put it in chronological order (as previously discussed). I re-classified Beral - author doesn't call it a "Meta-analysis" so we shouldn't either. No reason for Brind's opinion to change it around. I'm going back in to smooth out transitions and things, but please take a look at it and post what you think.

I am officially challenging the following sources (see examples in "CLINMED comments"):

• "Frezza" - this is a general repository of info with no reputation of fact-checking. It is not reliable. If someone can find the article on a reliable and verifiable site, like AIM's website, I will have fewer issues with it as a source of scientific critcism.
• "etters" - amateur partisan website with pro-life and religious bias, associated with conservative group known for hatemongering. No reputation for fact-checking. This one needs to go, period.
• "AnnieAppleseed.org" - this cites like one piece of info. If the info is legit and relevant, we can find it from a better source.

I am going to remove this material and post it here for possible resourcing.

The more debatable sources are Brind's self-publication sources and Malec's religious newsletter sources. The subject is covered in numerous reliable and verifiable sources. We have Brind's study. Given that the subject has been covered and well-documented in academia, and the page is not a soapbox for all of Brind's pro-life activism, why should we resort to unreliable sources? I think we need some discussion on this.

Consider these points from WP:Fringe theories:

• An appearance on Wikipedia should not make something more notable than it actually is. Since Wikipedia describes in its articles significant opinions, it is important that Wikipedia itself not become the notability-validating source for non-significant theories...Furthermore, one may not be able to write about a subject in a neutral manner if the subject completely lacks secondary sources that are reliable.
• Primary sources about research and investigations should only be used to verify the text and should not be relied upon exclusively as doing so would violate Wikipedia's ban on original research. In the case of obscure fringe theories, secondary sources that describe the theories should be carefully vetted for reliability. This includes references, citations, and external links.
• Inclusion and exclusion of content related to fringe theories and criticism of fringe theories may be done by means of a rough parity of sources. If an article is written about a well-known topic, fringe theories that may seem relevant but are only sourced by obscure texts that lack peer review should not be included in the article.
• Parity of sources may mean that certain obscure fringe theories are only reliably and verifiably reported or criticized in alternative venues from those that are typically used in publishing about these topics. For example, the lack of peer-reviewed criticism of creation science should not be used as a justification for marginalizing or removing scientific criticism of creation science since peer-reviewed journals routinely reject submissions relating to the subject.

Peer-review journals regular accept submissions relating to both abortion and breast cancer. The level of secondary reliable sources proves this. This theory has been thoroughly peer-reviewed. Sources of an unreliable nature should be discounted. This isn't creation science so religious sources aren't valid. It's not the Apollo hoax, so random amateur websites aren't valid. Let's just stick to WP standards. Thoughts? Phyesalis 17:55, 30 November 2007 (UTC)

I'm taking the lack of response as a lack of defense. I don't think there is an argument for keeping these sources. I'm going to remove the material and post it here so that if reliable and verifiable sources can be found, the info can be worked in. Phyesalis 05:05, 3 December 2007 (UTC)

Sounds fair. - RoyBoy ⁸⁰⁰ 04:21, 4 December 2007 (UTC) I agree amateur websites should be avoided for a Good Articles. The AnnieAppleSeed ref simply provide a plain language to a scientific studies findings. The findings can just as easily be ref'd by the study. - RoyBoy ⁸⁰⁰ 04:52, 4 December 2007 (UTC)

Removed content

Removed:

• Malec citation from catholiccitizens.org, changed "critics" to "Brind".
• The conservative political group Accuracy in Media (AIM) criticized the Lindefors-Harris study, claiming that the Lindefors-Harris control group was not well-defined and did not account for differences between how Swedish and American women use abortion. AIM also alleged bias in the study because funding came from Family Health International, a large pro-choice non-profit organization trying to meet the public health needs of the world's poorest people.
• "Scott Somerville of the conservative group Accuracy in Media claims that it took a long time for Howe's study to be published due to a number of American journals that rejected the article. The Howe study was published in the Britain-based International Journal of Epidemiology in 1989." Phyesalis 05:32, 3 December 2007 (UTC)

• Dr. Brind argues some interview-based studies indicate a statistically significant correlation, but debate remains as to the reliability of these retrospective studies because of possible response bias. Phyesalis 01:35, 4 December 2007 (UTC)

Round two

• added intro content to "Politicization" from Jasen.
• Changed "As a result, ABC is incorrectly referred to as pseudoscience" to "ABC is often referred to as 'pseudoscience'". Sources do not support "incorrectly", they support the appropriateness of the term.
• removed Jasen line after "pseudoscience" because her whole article is about how Brind's theory is a part of pro-life politics and how it was "discredited". Phyesalis 07:46, 3 December 2007 (UTC)

Okay, where do you get often from? Seems two steps too far. Pro-choice advocates (pro-lifers would call them zealots) call ABC pseudoscience. It is not a mainstream view, and I am getting a sense of deja-vu. - RoyBoy ⁸⁰⁰ 04:20, 4 December 2007 (UTC)

Me too! It's funny, because once again, it seems like sources have been used out of context to support ABC. The sources cited for "incorrectly" labeling ABC as bad science do not support the OR assertion that "pseudoscience" is an incorrect label. They (Mooney, Jasen, Weed and Kramer, the PublicEye article, the House Committee for Oversight and Governmental Reform report, the head of Epidemiology of the ACS, the NBCC Golden Boob award) all support (though some state outright) that ABC (a causative relationship) is bad, extremely bad, science. They use terms like "pseudoscience", "biased agenda driven science"; they assert that a causative link is a "leap beyond the bounds of inference" and that it is settled science (as in "purely a political debate"). None of them (that I recall) say that that ABC is legitimate science. Phyesalis (talk) 22:03, 4 December 2007 (UTC)

I'm still on a chocolately high, I'm soooo friggin mellow. I don't challenge your sources (yet :"); but I am questioning your application. At least some of Brind's assertions and pro-life interpretation of evidence are pseudoscience (or put more accurately, incorrect) particularly on causation; however we have to be very careful in applying that to the entire ABC issue and the ABC hypothesis mechanism. For example, I have found no one calling Russo and Russo's discoveries in rats pseudoscientific. I'm referring to DNA repair, cell cycle length and immature mammary cells being more vulnerable to carcinogens.

I think its preferable and more NPOV give such determinations specific context; and to ensure they do not come from amateur sources. I have always wanted a pseudo-science reference in the article and I actively sought one out (lawyer, then another - golden boob - for specific narrative reasons). With context, I added it as part of the North Dakota lawsuit. In doing so I made certain it was clarified who was saying it and when; and I made sure it was offset within the section for broader context; such as with the judges comment that the issue was in a "state of flux". This helps avoid giving the impression scientists "reject" or find the hypothesis itself pseudoscience. I should emphasize the pro-choice lawyer had no issue rendering an opinion that was contrary to the judges conservative and well considered interpretation.

Likewise, pro-choice sources you come across can equally be certain of their opinions, but cannot be bothered to differentiate between the lies of pro-life advocates and the findings of science. Its easy, convenient, even preferable to call it all pseudoscience. That's sociology rather than science rendering a verdict. *eats more chocolate to boost happy endorphins* - RoyBoy ⁸⁰⁰ 00:15, 5 December 2007 (UTC)

My two cents are that it's better to avoid the word "pseudoscience" wherever possible. Even in cases which are clearly considered pseudoscientific by the community, it's difficult or impossible to source adequately, and always makes people see red. Better to just explain clearly that the medical and scientific communities reject ABC, and that at present its support is restricted to pro-life advocacy groups. MastCell ^Talk 21:31, 12 December 2007 (UTC)