EXAMPLES OF BIAS IN THE SCIENTIFIC COMMUNITY

Misrepresentation of one's data is a form of scientific misconduct. A conventional definition of misrepresentation is to "omit a fact so that what is stated or presented as a whole states or represents a material falsehood."

In each of the following studies, the overall data offered by the scientists demonstrated that induced abortion was coupled with an increased risk of breast cancer; yet one would never have known this by reading the scientists' conclusions. Here are some examples of this kind of scientific misconduct:

1. "perhaps all that can be definitively stated is that any risk associated with induced abortion is at most statistically marginal." [1]

2. "Thus, our results indicate a lack of association between induced and spontaneous abortions and breast cancer risk." [2]

3. "Although we got the data described above, we are rather hesitant at the present moment in inducing some definite conclusions from them." [3]

4. "These data suggest that the risk of breast cancer is not materially affected by abortion, regardless of whether it occurs before or after the first term birth." [4]

5. "our data suggest that the risk of breast cancer associated with any pregnancy termination is likely to be small, if it exists at all." [5]

Another tactic employed by scientists is to offer no overall data in one's study concerning induced abortion and increased risk of breast cancer, but to include a statement within the conclusions that no positive association between abortion and breast cancer was found. Some examples include the following statements:

1. "The results are entirely reassuring ... there is no indication of any association between abortion before first term pregnancy and breast cancer risk." [6]

2. "a history of stillbirth and abortion did not emerge as significant risk factors for breast cancer in our study." [7]

3. "After controlling for other variables the risk significantly decreased with abortion history." [8]

4. "history of ... artificial abortion was also not markedly different between the two groups (cases and controls). [9]

5. "there was no evidence that one type of abortion (spontaneous or induced) was particularly harmful." [10]

Other scientists have been found to have stashed significant data showing abortion to be linked with breast cancer "in the file drawer." A perfect example is the Australian study on breast cancer. [11] Rohan et al examined various risk factors for the disease, including dietary risk factors. Thorough reproductive histories were obtained from the women which included abortion history. However, when the study was originally published in the American Journal of Epidemiology in 1988, mention of its most significant risk factor -- abortion -- was omitted.
This did not come to light until seven years later when French researcher Nadine Andrieu et al published a small meta-analysis examining data from six studies and reported that "induced abortion had actually been found by Rohan et al to be the strongest (stronger even than family breast cancer history) and most significant risk factor for breast cancer in Australian women." Andrieu found a synergistic effect (meaning that the whole exceeds the value of the sum of the two effects) between family history and abortion. [12]

Dr. Joel Brind, president of the Breast Cancer Prevention Institute, had this to say about Rohan's omission: "It was the strongest risk factor they found. It was the only one that was clearly statistically significant. And this you don't do. This is not what you see in scientific research, ever. I've never seen it before, where the most significant finding in a study is specifically left out of a research paper .... And we hypothesize that there is more of it." [13]

As a matter of fact, there is more of it. Other studies finding a significant positive association between abortion and breast cancer faced multiple rejections by major medical journals, and this was not due to a lack of ability or prestige of the authors of the study. [14, 15] The Bu study done on mainland China was co-authored by noted epidemiologist, Janet Daling, and her colleagues at the Fred Hutchinson Cancer Research Center located in Seattle, Washington. [14] This study's disturbing results demonstrated that one abortion nearly tripled an individual's risk of breast cancer. Its abstract was published in the American Journal of Epidemiology in 1995, but publication of the full paper was rejected by three major medical journals.

Publication of the Howe study, the only statistically significant prospective study, took nine long years because it was rejected by several journals, including the American Journal of Epidemiology. Howe et al reported an increased risk of 90% among New York state women. [15]
Finally, scientists have also been known to write misleading abstracts suggesting that a link between abortion and breast cancer was not found, although their data reveal increased risk. [16] Others have published studies that are too small [16, 17, 18] to implicate induced abortion as a risk factor for breast cancer.

REFERENCES

[1] Lipworth et al (1995) Int J Cancer 61:181-4.

[2] Tavani et al (1996) Int J Cancer 65:401-5.

[3] Segi et al (1957) GANN 48 (Suppl):1-63.

[4] Rosenberg et al (1988) Am J Epidemiol 127:981-9.

[5] Newcomb et al (1996) JAMA 275:283-7.

[6] Vessey et al (1982) Br J Cancer 45:327-31.

[7] Rao et al (1994) Br J Cancer 70:129-32.

[8] Gandra et al (1993) Acta Med Portuguesa 6:129-33 (in Portu
guese).

[9] Lai et al (1996) Proc Natl Sci Council, ROC 20:71-7.

[10] Wu et al (1996) Br J Cancer 73:680-6.

[11] Rohan et al (1988) Am J Epidemiol 128:478-89.

[12] Andrieu et al (1995) Br J Cancer 72:744-51.

[13] Joel Brind, Ph.D., "Abortion and Breast Cancer," Lecture at En
deavour Forum Public Meeting, August 24, 1999, Malvern, Victoria, p.
16-17.

[14] Bu et al (1995) Am J Epidemiol 141:S85.

[15] Howe et al (1989) In J Epidemiol 18:300-4.

[16] Marcus et al (1999) Am J Pub Health 89:1244-7.

[17] Newcomb et al (2000) Cancer Causes Control 11:777-81.1.

[18] Lazovich et al (2000) Epidemiol 11:76-80.