The Breast Cancer Epidemic: Modeling and Forecasts Based on Abortion and Other Risk Factors
Journal of American Physicians and Surgeons Volume 12 Number 3 Fall 2007
Using national cancer registration data for female breast cancer incidence in eight European countries—England & Wales, Scotland, Northern Ireland, the Irish Republic, Sweden, the Czech Republic, Finland, and Denmark—for which there is also comprehensive data on abortion incidence, trends are examined and future trends predicted.
Seven reproductive risk factors are considered as possible explanatory variables. Induced abortion is found to be the best predictor, and fertility is also a useful predictor. Forecasts are made using a linear regression model with these explanatory variables.
Previous forecasts using the same model and incidence data for years through 1997 for England & Wales are compared with numbers of cancers observed in years from 1998–2004 in an Appendix.
The forecast predicted 100.5% of the cancers observed in 2003, and 97.5% of those observed in 2004.
The Challenge of Abortion for Epidemiologists in Female Breast Cancer Research Trends
It is difficult for epidemiologists to discover women’s abortion history. In any study the numbers of women who have had abortions may be under-reported.
National data on abortions in most countries tends to be deficient, with abortions under-reported. Official abortion statistics in the United States and France are known to understate the numbers of legal induced abortions.The countries considered in this study are believed to have nearly complete official abortion counts.
The long lag time for the development of breast cancer magnifies the problem. The average age of diagnosis is over 60, while most abortions and live births occur at ages under 30.
The modern increase in breast cancer incidence is obvious at ages over 45, and Figure 1 for England &Wales shows the increase is small below age 45.
Abortion did not become legal in most Western countries until the 1970s, and earlier abortions among older women are not recorded. Consequently, the older women, whose breast cancer incidence is known, have abortions not detectable by a longitudinal study, while the younger women, whose abortion history is known, tend to be too young to have experienced most of the modern increase in breast cancer.
Where the increased risk is apparent, even under age 40 in a study free of recall bias, there is an acknowledged need to extend the study to women older than 40. The long time lags, however, can be used to make long-term forecasts of cancer trends…
The increase in breast cancer incidence appears to be best explained by an increase in abortion rates, especially nulliparous abortions, and lower fertility.
And the social gradient, which is not explained by fertility, seems also attributable circumstantially to abortion. A linear regression model of successive birth cohorts of women with abortion and fertility as explanatory variables fitted to the cancer incidence up to 1977 has produced forecasts that have performed well in the years 1998–2004 in Great Britain (AppendixA).
The new forecasts for eight countries can be tested in the coming years.
Patrick S. Carroll, M.A., is Director of Research, Pension and Population
Research Institute (PAPRI), 35 Canonbury Road, London N1 2DG, UK
Acknowledgements: Particular thanks are due to the charities LIFE and The Medical Education Trust, which funded the research, to the national statistical offices and cancer registries, which provided the data, and to the statisticians who kindly gave advice. Figure 10 is reproduced from the publication with permission of the Cancer Surveillance Team, Information Services Division (ISD), NHS National Services, Scotland. Computing was done by Andrew Chan and Lee Young.Potential conflicts of interest: none disclosed.[The Breast Cancer Epidemic: Modeling and Forecasts Based on Abortion and Other Risk Factors, Patrick S. Carroll, M.A., Journal of American Physicians and Surgeons Volume 12 Number 3 Fall 2007, http://www.jpands.org/vol12no3/carroll.pdf]
[The Breast Cancer Epidemic: Modeling and Forecasts Based on Abortion and Other Risk Factors, Patrick S. Carroll, M.A., Journal of American Physicians and Surgeons Volume 12 Number 3 Fall 2007, http://www.jpands.org/vol12no3/carroll.pdf]
ABORTION PREDICTS BREAST CANCER BETTER THAN OTHER FACTORS, STUDY SHOWS. A new study published in the Journal of American Physicians and Surgeons 3Oct07 finds that abortion is the best predictor of whether women will contract breast cancer. Abortion also is a better indicator of future breast cancer issue than six other commonly used factors.
Patrick Carroll of the Pension and Population Research Institute in London conducted this new study and showed that countries with higher abortion rates, such as England & Wales, could expect a substantial increase in breast cancer incidence.
"Induced abortion is found to be the best predictor, and fertility is also a useful predictor," he writes. "The increase in breast cancer incidence appears to be best explained by an increase in abortion rates, especially nulliparous abortions, and lower fertility."
He found that, when abortion rates are low such as in Northern Ireland and the Irish Republic, a smaller increase is expected.
And in nations experiencing a decline in abortions, like Denmark and Finland, a similar decline in breast cancer is anticipated.
Carroll used the same mathematical model for a previous forecast of numbers of breast cancers in future years for England & Wales. He based the model on cancer data up to 1997 that has proved quite accurate for predicting cancers observed in years 1998 to 2004.
In four nations — England & Wales, Scotland, Finland and Denmark — he discovered a social gradient unlike that for other cancers. He found upper class and upwardly mobile women have more breast cancer than lower-income women.
Carroll suggests that the known preference for abortion in this class might explain the phenomenon. Women pursuing higher educations and professional careers often delay marriage and childbearing. Abortions before the birth of a first child are highly carcinogenic, he explained.
Carroll used national data from nations believed to have "nearly complete abortion counts." Therefore, his study is not affected by recall bias.
Karen Malec, president of the Coalition on Abortion/Breast Cancer, commented on the new study in a statement sent to LifeNews.com.
"It's time for scientists to admit publicly what they already acknowledge privately among themselves that abortion raises breast cancer risk," she said. Malec said such scientists need to "stop conducting flawed research to protect the medical establishment from massive medical practice lawsuits."
Related web sites: Read the new study at http://www.jpands.org/vol12no3/carroll.pdf
[4Oct07, Ertelt, LifeNews.com London, England]