Those following daily are beginning to get confortable with the jargon (I hope). For ease of newcomers following along , please consult the glossary of terms that I’ve written to make the terminology very understandable. Also, consult the post that explains the essential background.
Title: Pregnancy related protection against breast cancer depends on length of gestation
Journal: British Journal of Cancer (2002) 87: 289-290
Authors: LJ Vatten, PR Romundstad, D Trichopoulos, R Skjarven
Funding: Norwegian Cancer Society
So far we have seen research that has linked induced abortion with an increased risk of breast cancer. We have seen authors who stand behind their methods and results, but who attribute the increased link to a purported recall bias, which is claimed by the pro-abort researchers to be a function of these retrospective (self-report) studies. We have also seen one of the papers upon which these claims have been made, and have seen the appallingly sloppy and irresponsible statistics employed which simply strain credulity. Further, we have seen that spontaneous abortion (miscarriage) has no effect on breast cancer risk, as the estrogen levels in the mothers is abnormally low in these pregnancies. I have also been assailed by the usual suspects who claim that the retrospective studies are too small (a fair claim for a few, but not the majority).
Today we turn our attention to a paper employing a method touted by the pro-aborts: a prospective study. In this study National birth and cancer databases in Norway were employed to get an objective assessment of women from first birth in 1967 until the diagnosis of cancer, death from any cause, or the end of the follow-up period on December 31, 1997.
This study should please the pro-abort crowd for a couple of reasons. It’s prospective and it’s large, very large: 694,657 women.
Results: The study substantiated the data from some of the comparatively smaller studies we have thus far considered. For instance, it showed an increasing incidence of breast cancer with increasing age at first birth (95% CI):
Less than 20 years 1.0 (reference value)
20-24 years 1.10 (10% increased risk)
25-29 1.30 (30% increased risk)
30-34 1.48 (48% increased risk)
35+ 1.56 (56% increased risk)
So we see in a large prospective study the validation of the trends found in the smaller studies.
Now for the length of gestation in first pregnancy. As the length of gestation decreases, the risk of breast cancer increases (95% CI):
40 week gestation 1.0 (reference value)
37-39 weeks 1.08 (8% increased risk)
32-36 weeks 1.11 (11% increased risk)
less than 32 weeks 1.22 (22% increased risk)
So, from this short communication we see the validation of previous studies. We also see some major holes through which one could drive a truck.
In such a large data set, why didn’t the authors separate out spontaneous from induced abortions? They make an incredible claim that one study by Melbye shows no difference in breast cancer risk in women with either induced or spontaneous abortions. This, even though one of this study’s authors (Trichoupoulos from Harvard) published data to the contrary with 95% CI (We’ve already studied two of those papers here at Coming Home in the last week). So the authors lumped miscarriage in with induced abortion and still showed a 22% increased risk of breast cancer.
They could have, and should have, shown the data for induced and spontaneous abortions separately in this prospective study to put the issue of recall bias to bed, one way or the other. It would also be a good basis of comparison for those two categories’ results in the retrospective literature.
So why didn’t they?
The answer lies in the data reported. The data validate all other findings in the retrospective literature. The data also validate the proposed biological model, of placental lactogen maturing the breast cells in the last trimester.
Note even here the dishonesty built into the numbers, as week 32 (the earliest week reported) is in the middle of the third trimester. At this point, according to none other than Melbye, 90 % of the protective effect is already in place. Thus, the authors stratified the risk inherent in the remaining 10% of protective effect yet unachieved in the third trimester. They did nothing to show the loss of protective effect. In short, they lied.
A further masking of induced abortion’s effects is the lumping of miscarriage with induced abortion. As there are typically many more miscarriages than abortions, and since all earlier studies show no increased risk of BC with miscarriage, this too dilutes induced abortion’s real numbers.
Was this masking of the numbers regarding an ABC link intentional? I believe it was. The single greatest issue at stake in this paper was a refutation or validation of recall bias as an issue in the retrospective body of literature. The authors have all of the data before them and could re-crunch the numbers.
Taking the weak assumptions that underlie ‘recall bias’, the even weaker numbers (weak to the point of absurdity in an undergraduate statistics class) of the proposed evidence of recall bias by Rookus and van Leeuwen, the ham handed approach to an exquisite prospective data set of almost 700,000 women and the lack of population stratification to bolster the issue of recall bias, this scientist simply cannot believe that Vatten, et al. were so over-simplistic by any means other than intentional design.
This study validated every other trend in the body of retrospective literature. Even the 22% increased risk of BC in women whose pregnancies were interrupted in the middle of the protective effect period scream a validation of the 30-50% increased risks for induced abortions in the retrospective studies whose smaller numbers were not massaged.
Such recklessness constitutes a betrayal of women who are entitled to fully informed consent prior to a surgical procedure such as abortion. If any other surgical procedure raised women’s risk of breast cancer by 50% and the data were similarly massaged careers would end and lawsuits would abound.
When I was a graduate student we were taught that the only thing of value a Ph.D. has is his/her integrity, without which all else is naught. If these authors can’t be trusted to competently assess and report the data when millions of women’s lives hang in the balance, then what good are such people to society, to the scientific community? Is the next grant, the next paper, the next academic promotion, the adulation of pro-abort colleagues, the loss of one’s academic soul worth it?
One need only consider the epidemiologically marginal 50% increased risk of BC multiplied by the 1.8 BILLION women worldwide who have had abortions since 1960 (and the scores of millions more to come in the coming decade) to get the frightful numbers of women who have endured, and will endure the effects of such betrayal.
Scientific integrity matters.
Read Full Post »