The Authors Reply:
In 1997 we published a paper showing no association between induced abortion and subsequent breast cancer risk in a population-based cohort study based on national registries including 1.5 million women with 10,246 cases of breast cancer. 1 Recently we indicated the importance of adjusting for age at first birth in this study 2 as a response to a letter by Sharpe. 3 We argued that otherwise the biological effect of age at first birth would wrongly be attributed to induced abortion. 2 Brind et al state in the first paragraph of their letter that they concur with this conclusion, but they use the rest of the letter to repeat their critique of our abortion study, 4 a critique that we previously have commented on point by point. 5
They argue that we should have used the same approach as in our recent publication concerning pre-term birth and breast cancer risk. 6 They thereby disregard, however, the fact that in the pre-term study the focus is on the importance of gestational age at delivery among parous women only. The same approach is not applicable in the abortion study because it would imply studying the effect of induced abortion in a cohort including only women who have had an induced abortion. Therefore, to call the methodology used in the pre-term study “corrected” compared with the abortion study methodology is simply incorrect.
Their insistence on not adjusting for birth-cohort is just as wrong as insisting on not adjusting for age at first birth. Without adjustment for birth-cohort effects, differences in risk behavior according to birth-cohort might be ascribed incorrectly as an effect of induced abortion. If this is the type of nonspecific effects they would like to attribute to induced abortion, as mentioned in their first paragraph, we can only disagree. In other words, adjusting for birth-cohort does not underestimate the biological effect of induced abortion on breast cancer risk, but rather excludes the possibility of overestimation due to differences in risk factors other than induced abortion between birth-cohorts.
For these reasons we stand firm by our results showing no association between induced abortion and breast cancer risk.
Per Kragh Andersen
1. Melbye M, Wohlfahrt J, Olsen JH, Frisch M, Westergaard T, Helweg-Larsen K, Andersen PK. Induced abortion and the risk of breast cancer. N Engl J Med 1997; 336:81–85.
2. Melbye M, Wohlfahrt J, Andersen PK. Re: Adjustment for age at first birth in etiologic studies of breast cancer involving exposures that may affect age at first birth (letter). Epidemiology 1999; 10:467.
3. Sharpe C. Adjustment for age at first birth in etiologic studies of breast cancer involving exposures that may affect age at first birth (letter). Epidemiology 1999; 10:95.
4. Brind J, Chinchilli VM. Induced abortion and the risk of breast cancer (letter). N Engl J Med 1997; 336:1834.
5. Melbye M, Wohlfahrt J, Andersen PK. Re: Induced abortion and the risk of breast cancer (letter). N Engl J Med 1997; 336:1835.
6. Melbye M, Wohlfahrt J, Andersen AM, Westergaard T, Andersen PK. Preterm delivery and risk of breast cancer. Br J Cancer 1999; 80:609–613.