BREASTFEEDING

Abortions and Breast Cancer
ABC Subject:
Breastfeeding before age 20 reduces breast cancer risk by 80%
(new study in American Journal of Public Health)

Dear Pamela Marcus (PhD),

Hello and thank you for your recent study of adolescent reproductive breast cancer risks.1 Two factors you covered are breastfeeding and induced abortion. Women under age 20 who breastfed reduced their BC risk by 80% (95% CI=0.1-0.6).1 Only women under age 20 who carry at least one pregnancy to term will have the possibility of this breastfeeding benefit. Your results in the area of reduced abortion show small risk increases but your sample size is too small to achieve 95% confidence of increased risk. As you may be aware, at least eighteen (18) studies in the field of 'ABC' (Abortion-Breast-Cancer') have achieved statistical significance with the result:

   1. ONE study (Burany, 1979) reported that induced abortion DECREASED BC risk
   2. SEVENTEEN (17) studies reported that induced abortion INCREASED BC risk

I.e. Over 94% of the statistically significant 'ABC' studies reported increased BC risk (see appendix).

Recall Bias in 'ABC' studies If women with breast cancer are more likely than controls to report induced abortions that they never had, then this would imply that the apparent 'ABC' risk may not exist. Such a theory of women imagining abortions that actually did not occur, may appear to be preposterous and bogus science. However, this is the basis of a 1991 study that reported seven Swedish women with BC (and one control) claiming induced abortions for which there was no SWEDISH record. The researchers labeled this 'over reporting' (imagining non-existent induced abortions).2 That these women very likely had these abortions outside Sweden (or inside Sweden but the procedure was recorded as something else) appears not to have crossed the minds of the researchers in 1991. In your 1999 AJPH report you refer to this study to justify possible 'reporting error' (i.e. recall bias).1 In 1998 the researchers of the Swedish study conceded that their over-reporting theory has no basis in fact:

       "Some women therefore had induced abortions abroad
         or unrecorded terminations of pregnancy [in Sweden]
         We are not surprised to find some Swedish women
         confidently reporting having had induced abortions
         during the period 1966-1974 which are not recorded as
         legally induced abortions."3

These researchers have downgraded their 'ABC' 'recall bias' idea to the status of a "notion".3 Pamela Marcus, if you are aware of even one study that provides statistically significant proof of recall bias in 'ABC' studies, please provide that reference. One of the researchers that you reference in your AJPH article, Polly Newcomb, recently wrote:

     "however, reproductive histories tend to be reported with a high degree of validity"4

You are rightly concerned about how reproductive factors affect BC risk in black American women. In 1993 the Journal of the National Medical Association reported that black women over the age of 49 years who had at least one induced abortion had an increased breast cancer risk of 370% (95% CI=2.6-8.4).5 It would take a very hefty dose of 'recall bias' to explain this result. And as you know, in a rat study those who had induced abortions had a 1300% increased risk of breast cancer.6 In rat studies there is no 'recall bias' possibility, since the rats are not interviewed. But, if they were, I'm sure they would not over-report.

       REFERENCES  

1 Marcus P, et al., AJPH, 1999;89:1244-1247

2 Meirik O, et al., American J Epidemiology, 1991; 134:1003-1008  

3 Brind J, Meirik O, et al.,  J Epidemiology & Community Health, 1998;52:    209-212  

4 Newcomb P, et al., American J Epidemiology, 1999;150:174-182  

5 Laing AE, et al,  J National Medical Association, 1993;85:931-939  

6 Russo J, Russo I, American J Pathology, 1980;100:497-512

Mike Richmond, newscan@vcn.bc.ca

 

P.S. Appendix: The eighteen (18) statistically significant 'ABC' studies:

     1. Bu et al., 1995. American Journal of Epidemiology, 141:S85  
         
Induced abortion increases risk of breast cancer

      2. Burany B (1979), Gestational  characteristics  in women with
          breast cancer.  Jugosl  Ginekol  Opstet 19:237-247  (in Serbo- -
         
Croatian); induced abortion DECREASES  breast  cancer risk.

      3. Daling JR, et al. (1994), Risk of breast cancer among young women:
          Relationship  to induced  abortion.  J Natl Cancer Inst  86:1584-1592.
         Women  who have an induced abortion before  age 18 +150%
        
have an increased breast cancer risk of 150%.

      4. Daling JR et al. (1996), Risk of Breast Cancer among White Women
          following Induced Abortion, American J Epidemiology, 144:373-380.
          Women with no births (nulliparous) and at least one induced +100%
         
abortion have an increased BC risk of 100% (95% CI=1.2-3.3.)

      5. Ewertz M, Duffy  SW (1988),  Risk of breast cancer in relation to
          reproductive  factors  in Denmark.  Br J Cancer  58:99-104. INDUCED +285%
         
abortion  increases  BC risk by 285%.

      6. Howe  HL, et al. (1989),  Early abortion and breast cancer risk
          among  women under age 40.  Int J Epidemiology  18:300-304. INDUCED +90%
         
abortion  increased  BC risk by 90%.

      7. Laing  AE, et al. (1993),  Breast cancer risk factors in African-American  women:
         
The Howard University  Tumor Registry Experience.
          J Natl Med Assoc  85:931-939.
          
Black women of age 50 or older who had at least one INDUCED +370%
         
abortion  have an increased  BC risk of 370%.

      8. Le MG, et al. (1984), Oral contraceptive use and breast or
          cervical cancer: Preliminary results of a French case-control
          study.  In: Wolff J-P, Scott JS, eds.  Hormones and sexual  factors
          in human cancer etiology.  Amsterdam:  Elsevier, 139-147.  INDUCED
         abortion  increases  breast cancer risk.

      9. Lipworth  L, et al. (1995),  Abortion and risk of breast cancer:
          a case-control study in Greece.  Int J Cancer 61:181-184.  INDUCED
          abortion  before first full-term pregnancy  increases BC risk +106%    by 106%.

    10. Luporsi, 1988. British Journal of Cancer, 72:744-751.
          Induced abortion increases risk of breast cancer

     11 . Newcomb PA, et al. (1995),  Pregnancy termination in relation to
          risk of breast  cancer.  Am J Epidemiology 141:S54 (abstract #215). +23%
         
INDUCED  abortion increases BC risk by 23%.

     12. Nishiyama  F (1982),  The epidemiology  of breast cancer in
          Tokushima  Prefecture.  Shikoku  Ichi  38:333-343 (in Japanese). +
         
INDUCED  abortion  increases  BC risk.

    13. Palmer, 1996. American J Epidemiology, 1996;143:S32 ..  Induced
          abortion  increases  risk  of  breast  cancer for nulliparous +40%
         
(i.e. childless) women; the increase in risk was 40%.

    14. Rohan et al., 1988. American Journal of Epidemiology, 128:478-489
          Induced abortion increases BC risk.

     15. Rookus  MA (1995),  Breast  cancer risk after induced abortion:
           A Dutch  case-control  study.  Am J Epidemiological  141:S54 +90%
           
(abstract #214).  INDUCED abortion increases BC risk by 90%.

     16. Rosenberg et al., 1988. Breast Cancer in Relation to the
           Occurrence and Time of Induced and Spontaneous Abortion. Am.
          J Epidemiology, 127:981-989.  Induced abortion first occurring
          after the first term birth increased BC risk by 40% (95% CI= +40%  1.0-1.9)

     17. Segi M, et al. (1957),  An epidemiological  study  on cancer in Japan.
         
GANN 48 (Suppl): 1:63.  INDUCED abortion increases breast cancer risk.

     18. Tavani A et al. (1996). Abortion and Breast Cancer Risk.
          International J Cancer, 65:401-405.  Women under age 30 who abort +30%
         
a first pregnancy increase BC risk by about 30%.