Several lines of evidence have long suggested that sex hormones play a central role in the etiology of breast cancer. As noted in the section Age-Incidence Curve of Breast Cancer Risk, rates of breast cancer increase rapidly in the premenopausal years, but the rate of increase slows sharply at the time of menopause, when endogenous hormone levels decline rapidly. In addition, several reproductive variables that alter estrogen status affect risk of breast cancer; for example, early age at menarche and late age at menopause are associated with increased risk of breast cancer. After menopause, adipose tissue is the major source of estrogen, and obese postmenopausal women have both higher levels of endogenous estrogen and a higher risk of breast cancer. In animals, estrogens and progesterone promote mammary tumors. Also, hormonal manipulations, such as administration of antiestrogens (e.g., tamoxifen citrate), are useful in the treatment of breast cancer and reduce breast cancer incidence in high-risk women. Despite this strong body of evidence suggesting a central role of endogenous estrogen, studies directly relating blood or urinary hormone levels to risk of breast cancer have largely been inconsistent. Only with the availability of large prospective studies has a clearer picture emerged.

Methodologic Issues in Studies of Endogenous Hormones and Breast Cancer Risk
In contrast to clinical determinations, in which discerning grossly abnormal from normal hormone levels is the focus, epidemiologic studies are usually aimed at detecting modest differences within the normal range of levels. Considerable laboratory error has been reported in studies of assay reproducibility, and several hormones are measured quite poorly by some laboratories. Low reproducibility could result in failure to detect true (and important) exposure-disease associations. Varying sensitivities and specificities of different laboratory assays also have made comparison of results between studies difficult. For example, in studies of plasma estradiol levels, mean levels in control subjects have ranged from 9 pg per mL81 to 28 pg per mL. Although these differences may result in part from differences in the characteristics of study subjects (i.e., differences in adiposity), a substantial component is probably due to the use of varying laboratory methods.

Several hormones, particularly estrogens, fluctuate markedly over the menstrual cycle. In some early studies, hormone levels were measured in samples collected without regard to the menstrual cycle phase, thus adding considerable noise to the comparison of hormone levelsbetween breast cancer cases and controls. This noise could mask true associations or, because of chance differences in the distribution of cycle phase between cases and controls, could result in associations that do not truly exist. Wih better understanding of such methodologic issues, studies have begun to collect all samples at approximately the same time in the cycle, have matched on cycle day, or have carefully controlled for cycle day in the analysis - all appropriate strategies.

For both logistic and financial reasons, in most epidemiologic studies, only a single blood sample can be collected per study subject. Whether a single sample can reflect long-term hormone levels (generally the exposure of greatest etiologic interest) is therefore an important issue. In several studies, repeated blood samples were collected over a 1- to 3-year period in postmenopausal women, and the correlation between the samples was calculated. Overall, levels of steroid hormones were reasonably stable, with intraclass correlations ranging from 0.5 to 0.9. This level of reproducibility is similar to that found for other biological variables, such as blood pressure and serum cholesterol measurements, all parameters that are considered reasonably measured and that are consistent predictors of disease in epidemiologic studies. Data on premenopausal women are much more limited, although androgen levels have been noted to be reasonably correlated over a several-year period.

The complexity and expense of hormone assays, coupled with the need to collect urine or blood samples from study subjects (and timed samples in premenopausal subjects) have resulted in both a limited number of epidemiologic studies of these issues and small sample sizes. These factors, in conjunction with error in the laboratory assays, have likely contributed to the lack of consistent findings. A summary, which relies when possible predominantly on the larger studies and those using a prospective design, is provided in the sections below.

Estrogens
Estradiol, considered the most biologically active endogenous estrogen, circulates in blood either unbound (“free”) or bound to sex hormone-binding globulin or albumin. Free or bioavailable (free plus albumin-bound) estradiol is thought to be readily available to breast tissue and thus may be more strongly related to risk than total estradiol. Postmenopausally, estrone is the source of most circulating estradiol, and estrone sulfate is the most abundant circulating estrogen. Both normal and malignant breast cells have sulfatase and aromatase activity, so estrone and estrone sulfate could serve as a ready source of intracellular estradiol.

Among postmenopausal women, the most consistent finding is a positive relationship between total estradiol and risk of breast cancer. In a meta-analysis of six prospective studies, breast cancer patients had mean estradiol levels that were 15% higher (p <.001) than those of healthy controls. Similarly, among 16 case control studies, breast cancer patients were noted to have mean levels 24% higher than levels in controls, although substantial heterogeneity in results was noted among these studies.

Results of the largest prospective study have been published since this meta-analysis. Blood samples were collected from 32,826 women in the Nurses’ Health Study from 1989 to 1990; among postmenopausal women not using hormone replacement therapy at blood collection, 156 women were diagnosed with breast cancer after blood collection. Two controls were selected per case, matched on age, menopausal status, and time of blood collection. A significant positive association between plasma estradiol level and risk of breast cancer was seen (top versus bottom quartile comparisons: multivariate relative risk, 1.9; 95% confidence interval, 1.2 to 3.5). Among women with no prior use of postmenopausal hormone therapy (for whom the single plasma hormone measure would best reflect past hormone levels), the association was markedly stronger (relative risk, 3.8; 95% confidence interval, 1.6 to 8.7). The comparable relative risk for women with high levels of plasma estrone sulfate was 4.3 (95% confidence interval, 1.9 to 10.1).

Other estrogens and estrogen fractions have not been as thoroughly evaluated as total estradiol. Percent free estradiol has been assessed in four prospective studies; a positive relationship was noted in three but not in the fourth. Percent bioavailable estradiol was positively associated with risk in one study but not in a second. Higher estrone levels were significantly associated with increased breast cancer risk in several studies but not in all. Similarly, findings for estrone sulfate (evaluated in only three studies) have been somewhat inconsistent.

Overall, the epidemiologic data now provide strong evidence for an influence of plasma estrogens on breast cancerrisk in postmenopausal women based largely on the substantial evidence for total estradiol. However, the specific estrogens and estrogen fractions most related to risk, and the dose-response relationship, remain to be defined.

Data on premenopausal estrogen levels and breast cancer risk are more limited, in large part because of the complexities related to sampling during the menstrual cycle. Data from several case control studies, but not others, suggest that high levels of estradiol in premenopausal women increase the risk of breast cancer. In the largest prospective analysis to date, involving 61 cases, estradiol levels were 12% higher in women who developed breast cancer than in controls. A second prospective study, with 22 cases, reported a nonsignificant increased risk for women with higher estrone and estradiol levels in the follicular phase but not the luteal phase of the menstrual cycle, whereas a third study found no association. To date, few studies have evaluated free or bioavailable estradiol levels in premenopausal women.