Breast Cancer and the Pill: Other Questions

Chris Kahlenborn
(chapter nine)
Breast Cancer: Its link to Abortion
and the Birth Control Pill
Reproduced with Permission

Q-9A: Have there been any other large studies such as those described previously concerning OCP use prior to a first full-term pregnancy (FFTP)?

Yes, another large study of 755 women was done by Chilvers1, but it confined itself to studying women under the age of 36. It found that women who took OCPs for 4 or more years prior to their FFTP had at least a 44% increased risk of breast cancer.

Table 9A:
The Chilvers Study of Women under the Age of 36

AUTHOR	FINDINGS	SIZE OF STUDY	YEAR STUDIED
Chilvers [1] United Kingdom Study	44% increase in women > 4 years use prior to their FFTP	755 less than age 36	1/82 - 12/85

Q-9B: Have any other studies shown an increased risk in young women less than the age of 36?

Yes, Pike et al showed that women age 32 or less had a 2.25-fold breast cancer risk if they took OCPs for 4 or more years prior to their FFTP2.

Q-9C: Does OCP use exhibit the multiplier effect when other risk factors are involved.?

When two factors affect the risk of obtaining breast cancer, and a woman has both of these risk factors, she can calculate her estimated risk by multiplying the independent risk of both of these factors. Several examples demonstrate that oral contraceptives increase the risk of breast cancer when combined with another risk factor. Brinton et al noted: "Among women with an affected relative, use of oral contraceptives for 5 or more years was associated with a RR of 3.1 (0.7-13.6) compared to a RR of 1.9 (1.2-3.1) among those without such a family history" [3, p.832]. Ravnihar4 noted that women who have a family history of breast cancer and used OCPs had a 5.2-fold risk over women who had a positive family history and had no use of OCPs. Last, Pike et al [2] noted that women without benign breast disease who used OCPs for at least 49 months prior to their FFTP had a 1.69-fold increased risk of breast cancer, but women who took OCPs and did have a history of benign breast disease had an infinitely increased risk as compared to the "control" population.

Q-9D: What do the other studies (whose data come predominantly after 1980) that have limited themselves to evaluating women under the age of 45 show for women who took OCPs prior to their FFTP?

Chapter 8 addressed the data and results of the four largest retrospective studies known to date. Table 9B shows the results of all of the "case/control" retrospective studies, the bulk of whose data come primarily after 1980. Eighteen out of 20 studies show an increased trend or significant risk in the development of breast cancer from oral contraceptive use before a woman's FFTP. In fact, most of the studies show an increased risk of greater than 40%. This means that the breast cancer risk of OCP use prior to a FFTP is almost certainly higher than the conservative estimate of 40% given in Chapter 8.

Table 9B:
Breast Cancer Risk From Starting OCPS Before A FFTP In Women under The Age Of 45

Author Change	Percent Change	Year	Study Size (RR) or Odds Ratio (OR)	Findings
Brinton et al [3]	42% increase*	1990-1992	1648< 45 years old	1.42 OR (calculated)
Chie et al [5]	10% increase*	1993-1994	97 premenopausal	OR=1.1 (0.2-5.5)
Chilvers [1]	2-51% increase	1982-1985	755 less than 36	1.02 RR for 0-4 yrs; 1.51 for 4-8 yrs; 1.44> 8 yrs pFFTP
Clavel [6]	50% increase**	1983-1987	358 premenopausal	OR=1.5 (1.0-2.2)
Ewertz [7]	50% increase**	1983-1984	203 less than 40	1.50 (0.87-2.58)
Lee [8]	90% increase**	?1988-1990	about 100 premenopausal	1.9 (0.4-7.7)
McCredie et al [9]	20% decrease**	1992-1995	467 under 40	0.8 (0.6-1.0)
McPherson et al [10]	2-97% increase**	1980-1984	351<45	1.02 (0.5-1.9) 1-12 months pFFTP; 1.97 (1.0-3.8) for 1-4 yrs pFFTP; 2.59 (1.3-4.5) > 4 yrs pFFTP
Meirik et al [11]	0-100% increase	1984-1985	422 less than age 45	1.2 (0.8-1.7 0-3 yrs; 1.0 (0.6-1.7) 4-7 yrs; 2.0 (1.8-4.2) > 8 yrs pFFTP
Miller et al [12]	100% increase	1983-1986	407 less than 45	2.0 (1.2-3.6)
Olsson [13]	80-110% increase	1979-1985	174 premenopausal	1.8 (1.0-3.2) for 0- 3 yrs and 2.1 (1.1- 3.8) for 4-7 yrs prior to FFTP
Palmer [14]	60-220% increase	1977-1992	219 less than age 45 (black women)	3.2 (1.1-9.2) 0-3 yrs; 1.7 (0.5-5.5) 3-4 yrs; 1.6 (0.4- 6.1) > 5 yrs pFFTP
Paul et al [15]	20% decrease**	1983-1987	155 less than 45 (severe stack effect)	0.80 (0.59-1.1
Rohan et al [16]	93% increase**	1982-1984	113 premenopausal	1.93 (0.44-4.42) for > 19 months use prior to FFTP
Rookus et al [17]	250% increase***	1986-1989	132 less than 36	3.5 (p< 0.01) starting at or before age 19
Rosenberg et al [18]	88% increase*	1977-1992	1427 less than 45	OR = 1.88 (calculated)
Ursin et al [19]	36% decrease to 74% increase	1983-1988	742 less than age 40	OR= 0.64 (0.32- 1.27) 1-12 months; 1.74 (0.75-4.03) > 13 months use pFFTP
Weinstein et al [20]	59% increase	1984-1986	about 326 less than or equal to 49	OR= 1.59 (1.02-2.47)
White et al [21]	50% increase***	1983-1990	747 < than 45 (suffers from stack effect)	OR= 1.5 (1.1-2.2) for > 1 yr use within 5 yrs of menarche (in parous women
Wingo et al [22]	40% increase	12/80-82	2089 less than age 45	1.40 (0.9-2.4) ages 20-34; 1.40 (1.1-1.8) ages 35-44


* Calculated from raw data. This result reflects a trend toward an increased or decreased risk but does not attain statistical significance. * FFTP in every study except Rookus and White who measure early OCP use in alternative ways.

The only two studies that showed an overall negative relationship were that of Paul et al [15] and McCredie [9]. The former study suffered from a huge stack effect having more than 6 times as many "controls" compared to "cases" in the 20 to 29 year-old age group (0.1035% vs. 0.0168%). The latter study found that women who had one child were at higher risk for breast cancer than nulliparous women, which raises concerns about the entire study. In addition, McCredie et al used the unconventional tactic of employing a p level greater than 0.05 to compute the data for risk of OCP use prior to a FFTP.

Q-9E: What about those women who used OCPs for a greater number of years prior to their FFTP?

We have already noted that Romieu23 found a 72% increased risk [RR=1.72 (1.36-2.19)] in her meta-analysis from 1990 in women who took OCPs for 4 or more years before their FFTP. In addition, the Brinton study [3] is significant in that she allowed a longer latent period to pass and found a 210% increased risk of deueloping breast cancer for young women (ie, under the age of 35) who took OCPs for more than 10 years and began before the age of 18.

Q-9F: Do OCPs taken after a FFTP cause breast cancer in women under the age of 45, if we confine the studies to those that obtained most of their information after 1980?

We note in Table 9C that most studies whose data comes predominantly after 1980 still show some risk of developing breast cancer from OCP use after a FFTP but in general this does not appear to be as great as the risk of OCP use before a FFTP.

Table 9C:
Breast Cancer Risk From Using OCPS After A First Full-Term Pregnancy (FFTP)

Author	Percent Change	Year	Study Size	Findings (RR) or Odds Ratio (OR)
Brinton et al [3]	15% increase**	1990-1992	1648< 45 years old	OR = 1.15
Chie et al [5]	80% increase*	1993-1994	about 80 less than 45	OR=1.8 (0.7-5.5)
Chilvers [1]	23-97% increase	1984-1988	755 less than 36	1.23 for 0-4 years; 1.97 for > 8 yrs
Ewertz [7]	no change	1983-1984	203 less than 40	1.00 (0.58-1.72
Mcpherson et al [10]	29% decrease**	1980-1984	351 less than 45	OR = 0.71
Miller et al [12]	160% increase	1983-1986	407 less than 45	2.6 (1.5-4.5)
Palmer [14]	20-150% increase	1977-1992	524 less than 45 (black women)	1.2 (0.6-2.1) < 3 yrs; 2.5 (1.2-5.3) 3-4 yrs; 1.7 (1.0-2.9) > 5 yrs
Rosenberg et al [18]	6% increase**	1977-1992	1427 less than 45	OR = 1.06
Ursin et al [19]	30% decrease to 23% increase	1983-1988	742 less than age 40	OR= 0.70 (0.47-1.03) 1-48 months; 0.89 (0.54-1.47) 4-8 yrs; 1.23 (0.63-2.40) > 8 yrs
Weinstein et al [20]	29% increase*	1984-1986	about 326 less than or equal to 49	1.29 (0.89-1.88)
White et al [21]	0-17% increase*	1983-1990	747 less than 45	OR = 1.00 (0.73-1.37) to 1.17 (0.85-1.61)
Wingo et al [22] CASH Study	0-40% increase*	12/80-82	2089 less than 45	1.40 (0.8-2.4) in age 20-34; 1.0 (0.8-1.3) in age 35-44


* This result reflects a trend toward an increased or decreased risk but does not attain statistical significance. ** OR calculated from raw data; see end of this chapter for details.

Q-9G: How does the risk of taking OCPs before a FFTP compare to taking them after a FFTP?

Table 9D compares the risks of OCP use before a FFTP to after a FFTP. One will note that in general, taking OCPs prior to a FFTP carries a higher risk than taking them after a FFTP.

Table 9D:
Comparison of Risks of developing Breast Cancer in Women Who Took OCPS Before or After Their FFTP

AUTHOR	YEAR	FINDINGS FOR RISK BEFORE FFTP	FINDINGS FOR RISK AFTER FFTP
Brinton et al [3]	1990-1992	42% increase	15% increase
Chie et al [5]	1993-1994	10% increase*	80% increase*
Chilvers [1]	1982-1985	2-51% increase	23-97% increase
Ewertz [7]	1983-1984	50% increase*	no change
Mcpherson et al [10]	1980-1984	2-97% increase*	29% decrease*
Miller et al [12]	1983-1986	100% increase	160% increase
Palmer [14]	1977-1992	60-220% increase*	20-150% increase*
Rosenberg et al [18]	1977-1992	88% increase	6% increase
Ursin et al [19]	1983-1988	36% decrease to 74% increase*	30% decrease to 23% increase*
Weinstein et al [20]	1984-1986	59% increase	29% increase*
White et al [21]	1983-1990	50% increase	0-17% increase*
Wingo [22] CASH Study	12/80-82	40% increase	0-40% increase


* This result reflects a trend toward an increased or decreased risk but does not attain statistical significance.

Q-9H: Do OCPs "as a whole," that is, when we compare "ever users" to "never users," cause breast cancer in women under the age of 45, when we confine the studies to those which obtained the bulk of their information after 1980?

We can see that the CASH study [22] found a 40% increased risk in women aged 20 to 34. Palmer [14], who studied young black women, found a 120% increased risk. Overall, Table 9E demonstrates clearly that the bulk of the studies since 1980, especially some of the larger ones, point to an increased risk for "ever" versus "never use."

Table 9E:
Breast Cancer Risk From Of Ever vs Never Use Of OCPS In Women under The Age Of 45

Author	Percent Change	Year	Study Size (RR) or Odds Ratio (OR)	Findings (RR) or Odds Ratio (OR)
Brinton et al [3]	27% increase*	1990-1992	1648< 45 years old	1.27 (1.1-1.5)
Chie et al [5]	60% increase*	1993-1994	about 80 less than 45	1.6 (0.7-3.8)
Chilvers [1]	6% decrease to 57% increase	1984-1988	755 less than 36	0.94 for 0-4 yrs; 1.42 for 4-8 yrs; 1.57 for > 8 yrs
Clavel [6]	50% increase	1983-1987	278 premenopausal	1.5 (1.1-2.1)
Ewertz [7]	18% increase*	1983-1984	203 less than 40	1.18 (0.73-1.91)
Gomes [24]	81% increase	1978-1987	71 < 43	1.81 (1.15-2.1)
La Vecchia [25]	30-90% increase	1991-1993	454 less than 45	OR = 1.3-1.9
Lee [8]	10% increase	?1988-1990	about 100 premenopausal	0.9 (0.5-1.6)
Lee et al [26]	13% increase	1982-1984	64 less than age 44	OR = 1.13 Costa Rican women
Lund et al [27]	29% increase	1984-1985	422 less than 45	OR = 1.29 (calculated)
Mayberry [28]	40-100% increase	1980-1982	177 less than 40 (all black)	1.4 (1.0-1.9) < 5 years; 2.0 (1.1-3.6) > 5 years
McCredie et al [9]	4% increase	1992-1995	467 under 40	OR = 1.04
Miller et al [12]	100% increase	1983-1986	407 less than 45	2.0 (1.4-2.9)
Newcomb [29]	0-40% increase*	1988-1991	1050 < 45	1.4 (0.8-2.3) in women < 35; 1.0 (0.8-1.3) 35-44
Noonan et al [30]	26% increase	1979-1984	301 less than 35	1.26 (0.95- 1.66) severe stack effect
Palmer [14]	120% increase	1977-1992	184 less than 45 (black women)	2.2 (1.5-3.3)
Paul et al [15]	no change	1983-1987	489 premenopausal	1.0 (0.77-1.3) severe stack effect
Primic-Zakelj et al [31]	4% increase*	1988-1990	501 premenopausal	1.04 (0.80-1.37)
Rosenberg [32]	76% increase	1982-1986	79 less than 40 (Canadian women)	OR = 1.76
Rosenberg et al [18]	10% decrease to 70% increase	1977-1992	1427 less than 45	1.7 (1.3-2.3) ages 25-34; 0.9 (0.7-1.0) ages 35-44
Ursin et al [19]	17% decrease	1983-1988	742 less than age 40	OR = 0.83 (0.62-1.12)
Weinstein et al [20]	68% increase	1984-1986	about 326 less than or equal to 49	OR= 1.68 (1.16-2.42)
White et al [21]	2% increase*	1983-1990	747 < than 45	OR = 1.02 (0.71-1.48)
Wingo [22] CASH Study	10-40% increase*	12/80-82	2089 less than age 45	1.4 (1.1-2.1) in age 20-34; 1.1 (0.9-1.3) in ages 35-44


* This result reflects a trend toward an increased or decreased risk but does not attain statistical significance.

Next Page: Q-9I: Why does Table 9E exhibit that ....
1, 2, 3