Harvard Reports on Cancer Prevention
Volume I: Human Causes of Cancer
Cancer Causes & Control:
An International Journal of Studies of Cancer in Human Populations
Official Journal of the International Association of Cancer Registries
Volume 7 Supplement November 1996 ISSN 0957-5243

Reproductive Factors

Introduction
Reproductive factors have been evaluated in relation to several types of cancer, including breast, ovary, endometrium, cervix, and colon cancers. Reproductive factors generally are considered to be characteristics that are related specifically to reproduction, such as number of births (parity), age at first birth, and lactation (nursing), or that mark a change in a woman's reproductive capacity, such as age at menarche and age at menopause. Each of these factors is associated with substantial hormonal changes. Menarche marks the beginning of menstruation and ovulation, and each pregnancy is associated with large increases in circulating hormones, particularly estrogens. Menopause marks the end of menstruation and is associated with large decreases in estrogen levels and increases in hormone levels compared with premenopausal women. It is because of these hormonal changes that reproductive factors have been hypothesized to influence cancer risk.

Breast cancer
Nulliparity (having no children), late age at first birth, early age at menarche, and late age at menopause all have been associated consistently with an increase in breast cancer risk.^1,2 For each of these factors, breast cancer risk tends to increase incrementally throughout the range of the variable so that there is no single high risk or low risk group. For example, risk tends to rise with increasing age at menopause from before 40 years of age to after 50 years of age. Findings for the relationship between lactation and breast cancer risk have been much less consistent; generally, no association or an inverse association (primarily among premenopausal women³) has been reported.

In general, reproductive factors are related modestly to risk of breast cancer in women. For example, women who have a first birth after age 30 years have a 50 to 100 percent higher risk of breast cancer relative to those who had their first child by age 20. Women who reached menarche by age 12 have only a 20 to 30 percent higher breast cancer risk compared with women who reached menarche at age 14 years.

The relationship between parity and breast cancer risk is more complex. Relative to nulliparous women, breast cancer risk actually is increased for one to two decades after giving birth, perhaps because of the increased exposure to circulating steroid hormones during pregnancy.⁴ After this time, however, breast cancer risk is lower in parous women compared with nulliparous women. This delayed (but long-lasting) reduction in risk may be related to hormone-induced changes in the cells of the breast, which result in their decreased susceptibility to carcinogens.⁵ Overall, the reduction in risk associated with parity outweighs the initial increase in risk, as the reduction occurs later in life when a woman's absolute breast cancer risk is much higher.

Although these associations are modest, reproductive factors are thought to account for a substantial part of the difference in breast cancer rates observed internationally, because the difference among countries for each of these factors is larger than that generally observed within a single country. For example, the average age at menarche is 12 or 13 years in the United States and other developed countries compared with a mean age of 17 years in rural China.

The precise mechanisms underlying these associations are as yet unknown. One or more estrogens or other growth factors are hypothesized to increase risk by promoting cell division and thereby increasing the possibility of genetic errors.^6-7 An early age at menarche and a late age at menopause would result in a longer lifetime exposure to higher estrogen levels and thus a greater breast cancer risk. In addition, the hormonal changes associated with pregnancy may lead to changes in breast tissue, which over the long run actually decrease this tissue's susceptibility to carcinogens.⁵

Endometrial cancer
The incidence of endometrial cancer has been linked closely to exposure to unopposed estrogen (that is, estrogen in the absence of progesterone) and as a result is related to a number of reproductive variables.^8-10 Late age at menopause is associated positively with this disease, presumably due to the longer period of exposure to high premenopausal estrogen levels. Early age at menarche generally has been found to increase endometrial cancer risk, although whether this association is independent of obesity (a known risk factor for endometrial cancer) is unclear. As with breast cancer, parity appears to lower the risk of endometrial cancer. Only a handful of studies have evaluated the relationship between age at first birth and endometrial cancer, and their findings have been inconsistent and generally do not support any substantial association.

Ovarian cancer
Most ovarian cancers arise from epithelial cells that make up the external surface of the ovary. Ovulation may increase ovarian cancer risk by subjecting these epithelial cells to repeated minor trauma (occurring when the ovum breaks through the surface epithelium during ovulation), increased cellular division (associated with repair of the surface epithelium after each ovulation), or exposure to hormone-rich fluid that surrounds the ovum.¹¹ This theory suggests that factors, such as parity, that decrease the number of ovulations would decrease ovarian cancer risk. High levels of gonadotropins, hormones secreted by the pituitary gland, also have been hypothesized to increase ovarian cancer risk. Since pregnancy decreases gonadotropin levels and a late age at menopause postpones exposure to high postmenopausal gonadotropin levels, each of these factors would be expected to reduce ovarian cancer risk according to this hypothesis.

By far, the most consistent finding in epidemiologic studies of ovarian cancer has been the inverse association with parity.^12-14 A trend of increasing protection from this cancer with increasing parity has been observed in most studies. Generally, women who have had three or more children have a 30 to 50 percent lower risk of ovarian cancer relative to nulliparous women.

Results have been much less consistent for age at first birth, age at menarche, age at menopause, and lactation in relation to ovarian cancer risk. In a combined analysis of several European studies, nulliparous women and women having a first pregnancy after age 34 had a modestly higher risk of cancer compared with women who had their first child at an earlier age.¹³ In contrast, in a pooled analysis¹² and a large prospective study,¹⁴ little association was found. Given the evidence to date, age at first birth, at least in women aged less than 35 years, does not appear to have a substantial effect on ovarian cancer risk.

In most studies, little if any relationship has been observed between late age at menarche and ovarian cancer risk,^12,14,15 although in a case-control study conducted in China,¹⁶ an age at menarche at 18 or older was found to be protective. Overall, there appears to be little effect of age at menarche on ovarian cancer risk, at least within the range of ages that menarche typically occurs in Western societies. Similarly, no consistent association has been noted between age at menopause and ovarian cancer risk. In most studies, age at menopause generally has been found to be either unrelated or positively related to risk.

Cervical cancer
The reproductive risk factors most consistently associated with an increase in risk of cervical cancer are having had multiple sexual partners and having an early age at first intercourse,¹⁷ factors that are probably primarily surrogates for infection with human papilloma virus, a well-substantiated cause of cervical cancer. In several, but not all, studies where the number of sexual partners and age at intercourse could be accounted for in the analysis, high levels of parity were found independently to increase risk.^18,19 This association may be related to the hormonal changes of pregnancy or to cervical trauma occurring during childbirth. Neither age at menarche nor age at menopause have been associated with cervical cancer risk.

Colon cancer
An observation that nuns are at higher risk of colon cancer relative to other women—in addition to gender differences in colon cancer incidence rates—first sparked interest in the relationships between reproductive factors, particularly parity and colon cancer risk.²⁰ Results of epidemiologic studies have been inconsistent however, and although a modest association between parity and colon cancer cannot be ruled out, no substantial association is likely to exist.^20-23 Age at first birth also does not appear to influence colon cancer risk. Other risk factors, such as dietary factors and physical activity, are of greater importance to colon cancer incidence in addition to being more amenable to public health intervention.

Other factors related to reproduction
Several other reproductive factors have been hypothesized to increase or decrease risk of specific cancers. Menstrual cycle length and regularity may be associated with risk of breast, endometrial, and ovarian cancers, although relatively few studies have addressed these relationships, and consistent findings have yet to emerge. Tubal ligation has been associated consistently with a decrease in risk of ovarian cancer, although estimations of the magnitude of this reduction varies substantially among studies. In several, but not all, studies, men who have had a vasectomy had an increased risk of prostate cancer of 30 to 100 percent compared with men who did not have this procedure. This increased risk generally was observed only after a period of 10 to 20 years.

Conclusion
Study results regarding the association between reproductive factors and cancer vary a great deal by type of cancer. For instance, nulliparity, late age at first birth, early age at menarche, and late age at menopause have been found consistently to be associated with increased breast cancer risk. Similarly, ovarian cancer has been found consistently to be associated with nulliparity. In contrast, there is no solid evidence of an association between reproductive factors and colon cancer. In some cases, reproductive factors are associated with a twofold increase in cancer risk, but, in general, the elevated risk is somewhat lower.

Summary Points

Table 1.

Suggestions

Suggested Further Reading

References