Pill Use Is Associated with Reductions in Overall Risk of Cancer and in Risk of Main Gynecologic Cancers
Ever-use of the pill had no adverse effect on the overall risk of cancer in the large cohort of British women participating in the Royal College of General Practitioners' oral contraception study. Rather, analyses of data reflecting as much as 36 years of observation indicate that ever-users of oral contraceptives had a 12% reduction in the risk of developing any cancer and a 29% reduction in the risk of developing cervical, uterine or ovarian cancer. (Analyses of data from a subset of the cohort, however, revealed no association between ever-use and the risk of any cancer.) Long-term pill use was associated with elevated risks of some cancers and with reduced risks of others.1 Analyses of data from the U.S. Nurses' Health Study, another long-term cohort study, confirm the inverse association between pill use and ovarian cancer risk; they also show that the risk of this disease is reduced among sterilized women and elevated among women who have used an IUD or are infertile.2
The British Study
The original British cohort comprised about 23,000 current pill users and a similar number of never-users recruited by general practitioners throughout the United Kingdom in 1968–1969. Participants were 29 years of age, on average, and were married or cohabiting at recruitment; most were white. They were followed up by their physicians, who collected information every six months about their pregnancies, illnesses, surgeries and use of hormonal contraceptives or hormone replacement therapy. One-quarter of women remained in the study until 1996. In addition, central registry data on cancer and mortality after the mid-1970s were available for three-quarters of the original cohort, regardless of whether the women were still being followed up by their physicians; these sources covered the period up to a woman's first cancer diagnosis or 2004, whichever came first.
Two data sets were used for the analysis of cancer risk. The main one was based on all women for whom central registry data were available and included about 744,000 woman-years of observation for ever-users of oral contraceptives and 339,000 woman-years of observation for never-users. The second one contained only information collected by general practitioners through 1996 and contained about 224,000 and 331,000 woman-years of observation for ever- and never-users, respectively. Researchers calculated the rates of first diagnoses of a variety of cancers among ever- and never-users; rates were standardized for women's age and parity at diagnosis, and for cigarette smoking and social class (as defined by husband's occupation) at recruitment. Relative risks were calculated to compare rates by use status and selected characteristics of women.
Participants included in the main data set were predominantly younger than 40 when they entered the study (94% of ever-users and 90% of never-users); most had had at least one birth (83% and 80%, respectively) and had husbands who were employed in manual occupations (64% and 61%). Close to half of ever-users and four in 10 never-users in this data set smoked. Some 13% of ever-users and 10% of never-users in the general practitioner data set had used hormone therapy.
Compared with never-users of oral contraceptives, ever-users in the main data set had a 12% lower risk of developing any cancer during follow-up and a 29% lower risk of developing one of the main gynecologic cancers (cervical, uterine or ovarian cancer). They had significantly reduced risks of cancer of the large bowel or rectum (relative risk, 0.7), uterus (0.6) and ovaries (0.5), and of cancers for which the site was unknown (0.6) or that were classified as "other" (0.9). The reduction in overall risk translates into an estimated 45 cancers prevented per 100,000 woman-years. Relative risks calculated from the general practitioner data set were significant only for uterine and ovarian cancer (0.5 for each). In the main data set, the overall risk of cancer was significantly reduced for ever-users of the pill who were in their 30s or 50s, among both smokers and nonsmokers, among women of most parities and regardless of social class.
Data from the general practitioners' observations were used to explore the relationship between cancer risk and characteristics of women's oral contraceptive use. These analyses showed no relationship between pill use for less than four years and cancer risk (median duration of use was 44 months), and a modest decrease in overall risk associated with use for 4–8 years (relative risk, 0.9). However, use for more than eight years, which accounted for less than a quarter of use in the cohort, was associated with an elevated risk of any cancer (1.2) and of cancers of the cervix (2.7) and the central nervous system or pituitary (5.5). Furthermore, the trend toward increased risks of these specific cancers with increasing duration of use was statistically significant, as was a trend toward decreased risks of uterine and ovarian cancer.
Ovarian cancer risk was reduced for up to 15 years after women had last used the pill, and uterine cancer risk was reduced for up to five years since last use; for both of these cancers, the data suggest continued reductions in risk at longer durations since last use. Although trends for individual cancers were not statistically significant, ever-users' risk of developing any main gynecologic cancer declined as the time elapsed since last use of the pill increased.
The researchers comment that "many women, especially those who used the first generation of oral contraceptives many years ago, are likely to be reassured by [these] results." Nevertheless, they acknowledge that their findings may not reflect current pill users' experiences, given changes in preparations and in use protocols. Moreover, they emphasize that "the likely balance of cancer risks and benefits" may vary in different parts of the world, and that this is an important area for further study.
The U.S. Study
The initial cohort of the Nurses' Health Study consisted of almost 122,000 married, female registered nurses who were 30–55 years old at recruitment, in 1976. Baseline data, including information on oral contraceptive use and risk factors for cancer, were collected in a mailed questionnaire. In follow-up questionnaires sent to participants twice a year, women were asked about their cancer risk factors and newly diagnosed diseases; follow-up continued through May 2004. Until the mid-1980s, follow-up questionnaires assessed contraceptive use among premenopausal women; 1994 questionnaires asked again about sterilization. In 1980 and 1992, women were asked about infertility.
The 28 years of follow-up yielded information on 2.5 million woman-years of experience, including data on 612 women who developed ovarian cancer and for whom duration of pill use was known. Researchers examined ovarian cancer risk in relation to duration of pill use and time since last use in analyses controlling for age, body mass index, parity, age at menopause, duration of postmenopausal hormone use, and history of sterilization and smoking. They found that risk declined significantly with increasing duration of pill use; women who had taken oral contraceptives for more than 10 years were less likely than never-users to develop the disease (relative risk, 0.6). The relationship between risk and time elapsed since last use did not demonstrate a significant trend, but women who had last used the pill 5–10 years earlier had a reduced risk of ovarian cancer (0.5). Risk also was reduced for those who had used oral contraceptives for more than five years and had last taken the pill within the past 20 years (0.6); no protective effect was seen for women who had discontinued use longer ago. The researchers speculate that the "waning" of the protective effect of taking the pill could be problematic, since the incidence of ovarian cancer is highest after menopause.
Use of two other contraceptive methods also was associated with the risk of ovarian cancer. Women who had undergone tubal li-gation had a reduced risk of this disease (relative risk, 0.7), and ever-users of an IUD had an elevated risk (1.8). Women reporting a history of infertility also had a somewhat elevated risk (1.4). The mechanisms underlying these associations are not well understood, as the researchers note, and require further investigation.—D. Hollander
1. Hannaford PC et al., Cancer risk among users of oral contraceptives: cohort data from the Royal College of General Practitioner's oral contraception study, BMJ, 335(7621):651–658.
2. Tworoger SS et al., Association of oral contraceptive use, other contraceptive methods, and infertility with ovarian cancer risk, American Journal of Epidemiology, 2007, 166(8):894–901.