Bone Mass and Breast Cancer Risk in Older Women: Differences by Stage at Diagnosis

Department of Epidemiology, Graduate School of Public Health, University of Pittsburgh, PA 15261, USA.
JNCI Journal of the National Cancer Institute (Impact Factor: 12.58). 07/2001; 93(12):930-6. DOI: 10.1093/jnci/93.12.930
Source: PubMed


Older women with low bone mineral density (BMD) have a decreased incidence of breast cancer. It is not known whether this association is confined to early-stage, slow-growing tumors.
We prospectively studied 8905 women who were 65 years of age or older during the period from 1986 through 1988 and had no history of breast cancer. At study entry, we used single-photon absorptiometry to measure each woman's BMD at three skeletal sites: the wrist, forearm, and heel. The women were followed for a mean of 6.5 years for the occurrence of breast cancer. All statistical tests were two-sided.
During 57 516 person-years of follow-up, 315 women developed primary invasive or in situ breast cancer. Multivariate analyses that adjusted for age, obesity, and other covariates revealed that the risk of breast cancer for women in the highest quartile of BMD for all three skeletal sites was 2.7 (95% confidence interval [CI] = 1.4 to 5.3) times greater than that for women in the lowest quartile at all three skeletal sites. The magnitude of increased risk associated with high BMD differed by the stage of disease at diagnosis and was greater for more advanced tumors (relative risk [RR] for TNM [i.e., tumor-lymph node-metastasis] stage II or higher tumors = 5.6; 95% CI = 1.2 to 27.4) than for early-stage disease (RR for in situ/TNM stage I tumors = 2.2; 95% CI = 1.0 to 4.8).
Elderly women with high BMD have an increased risk of breast cancer, especially advanced cancer, compared with women with low BMD. These findings suggest an association between osteoporosis and invasive breast cancer, two of the most prevalent conditions affecting an older woman's health.

10 Reads
  • [Show abstract] [Hide abstract]
    ABSTRACT: This paper introduces TRSkit, a simple and effective toolkit for building digital libraries on the World Wide Web. The toolkit was developed for the creation of the Langley Technical Report Server and the NASA Technical Report Server, but is applicable to most simple distribution paradigms. TRSkit contains a handful of freely available software components designed to be run under the UNIX operating system and served via the World Wide Web. The intended customer is the person that must continuously and synchronously distribute anywhere from 100 - 100,000's of information units and does not have extensive resources to devote to the problem. Journal of Internet Cataloging, 1(2), 1997, pp. 41-55. TRSkit: A Simple Digital Library Toolkit Michael L. Nelson <> Sandra L. Esler <> NASA Langley Research Center MS 157A Hampton, VA 23681-0001 January 31, 1997 Abstract This paper introduces TRSkit, a simple and effective toolkit for building digita...
    Journal of Library Metadata 11/1997; 1. DOI:10.1080/19386389709512376
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: The purpose of this study is to investigate the relationship between breast cancer risk, bone mass, and two polymorphic hormone receptor genes-- the estrogen receptor (ER) and vitamin D receptor (VDR) genes. We also planned to explore a possible functional mechanism to explain this association. Our target was to recruit 200 new breast cancer cases and 200 controls, ages 40-85, with equal numbers of African-Americans and whites. Bone mineral density (BMD) measurements of the forearm were to be obtained. We have enrolled 231 cases and 198 controls, with an age range of 39-84 years. There is an equitable distribution of the two ethnic groups, with 50.3% of the cohort being white and 49.7% African-American. We enrolled more than 400 subjects because some individuals changed their minds about giving blood, some blood samples were not analyzable, and some subjects have no bone density data due to instrument malfunctions. Genotype frequencies of the VDR and ER gene segments that we investigated were not significantly different in cases or controls. However BMD (expressed as a z-score) in the proximal radius of the cases is significantly higher than controls, as hypothesized. Furthermore, the odds ratio associated with having higher than average bone density was 1.98 (95% C.I. 1.32-2.97). This is our most significant and clinically relevant finding. Adjustment for potential confounders did not appreciably change this odds ratio. We are in the process of obtaining tumor tissue samples from subjects who are homozygous and who have given consent to our using their tissue for our final objective--to investigate the responsivity to estrogen of the polymorphic ER genotypes.
  • [Show abstract] [Hide abstract]
    ABSTRACT: Breast cancer risk reduction now represents an achievable medical objective. Current interventions include selective estrogen receptor modulators (SERMs), prophylactic surgery, and lifestyle change. For SERMs, current evidence supports tamoxifen use for breast cancer risk reduction whereas raloxifene requires further study. Prophylactic mastectomy and prophylactic oophorectomy, effective in retrospective clinical experiences, should be considered only for women at substantial risk willing to accept the irreversible consequences of these procedures. Although dietary fat intake is under clinical trial evaluation, lifestyle change, including weight loss, dietary change, and increased physical activity, can be recommended based on other health considerations. Use of any intervention requires careful breast cancer risk assessment, risk-benefit calculations, and informed decision making with full patient participation. Future breast cancer risk assessment may incorporate additional biologic measures of estrogen exposure and/or analyses of collected breast cells. Under active evaluation are novel SERMs, aromatase inhibitors/inactivators, gonadotrophin-releasing hormone agonists, retinoids, statins, and tyrosine kinase and cyclooxygenase-2 inhibitors.
    Annual Review of Medicine 02/2002; 53(1):519-40. DOI:10.1146/ · 12.93 Impact Factor
Show more


10 Reads