Bone Density and the Risk of Breast Cancer in Postmenopausal Women


Synopsis: The risk of breast cancer correlates with bone mass.

Source: Zhang Y, et al. N Engl J Med 1997;336:611-617.

Zhang et al from the framingham study studied the association of the incidence of breast cancer with metacarpal bone mass in a cohort of 1373 women. These women had bone mass assessed by hand radiography and measurement of the cortical width in the second metacarpal bone between 1967 and 1970. In the cohort, 91 women had developed breast cancer by 1993. Relative risk for breast cancer was increased, achieving statistical significance, in the group with the highest quartile of bone mass. In this group with 44 cases of breast cancer, the relative risk was 3.5 (confidence interval, 1.8-6.8). The authors conclude that this association between high bone mass and the risk of breast cancer was consistent with greater cumulative exposure to estrogen.


We have long been aware that there is a correlation between risk of breast cancer and conditions that are characterized by greater cumulative exposure to estrogen. Examples include an increased risk of breast cancer with an earlier age of menarche, a later age of menopause, obesity during the premenopausal years, and possibly anovulation. This study agrees with that reported in JAMA recently where the risk of breast cancer was 2.0-2.5 times higher comparing women with the highest quartile of bone density to the women with the lowest quartile (Cauley JA, et al. JAMA 1996;276:1404). These reports are consistent, therefore, with the general notion that a woman’s lifetime exposure to estrogen is associated with a slightly increased risk of breast cancer.

It is hard to quarrel with the conclusion that estrogen influences the risk of breast cancer. However, that is not the same thing as saying that estrogen causes breast cancer. Even though I believe this conclusion is consistent with everything we have known over the last 30 years, I can still find fault with this particular study. For example, women with the highest bone densities who showed the greatest risk of breast cancer were different. Overall, they had fewer children. Remember that pregnancy and multiple pregnancies, especially early in life, protect against breast cancer. They had a higher bone body mass index, consumed more alcohol, and smoked more. Of course, the epidemiologists in the study report that they controlled for these variables by mathematical manipulations. It is hard to know how reliable these manipulations are and whether they truly correct for the fact that these women with higher bone density had a collection of factors known to be associated with an increased risk of breast cancer. Furthermore, the total number of cases, 44, is not large; this is reflected in the relatively wide spread in the confidence interval.

Despite this criticism, I believe the preponderance of evidence is consistent with the fact that the cumulative exposure of a woman to estrogen during her reproductive years influences her risk of breast cancer. That does not mean that long-term exposure to postmenopausal hormone therapy necessarily increases one’s risk of breast cancer. The state-of-the-art with postmenopausal hormone therapy continues to be inconsistency and lack of agreement among more than 40 case-control and cohort studies. I continue to believe that this lack of agreement is good news, indicating that there cannot be a major impact. Otherwise, most, if not all, of the studies would be telling us the same thing.