Current issues of ACP Journal Club are published in Annals of Internal Medicine


Review: Recent HRT increases the risk for breast cancer

ACP J Club. 1998 May-June;128:73. doi:10.7326/ACPJC-1998-128-3-073

Source Citation

Collaborative Group on Hormonal Factors in Breast Cancer. Breast cancer and hormone replacement therapy: collaborative reanalysis of data from 51 epidemiological studies of 52 705 women with breast cancer and 108 411 women without breast cancer. Lancet. 1997 Oct 11;350:1047-59.



To determine, using meta-analysis, whether an association exists between hormone replacement therapy (HRT) and risk for breast cancer.

Data sources

Studies were identified from review articles, literature searches, and colleagues.

Study selection

Studies were selected if they included ≥ 100 women with breast cancer and if information was obtained on HRT use and reproductive and menopausal history.

Data extraction

Data on individual women were sought for use of HRT and hormonal contraceptives, sociodemographic factors, family history of breast cancer, height, weight, age at menarche, reproductive history, gynecologic surgery, menopausal status, age at menopause, and tumor spread (for those who had breast cancer).

Main results

51 studies from 21 countries involving 52 705 women with invasive breast cancer and 108 411 women without breast cancer were included (81% of eligible studies). Analyses were stratified by study; center within study; age at diagnosis; time since menopause; body mass index; parity; and, when appropriate, age of the woman when her first child was born. The main analyses were based on 53 865 postmenopausal women (17 949 with breast cancer and 35 916 without breast cancer) for whom age at menopause was known. The overall median age at first use of HRT was 48 years. The relative risk (RR) for breast cancer among women who had ever used HRT compared with those who had never used HRT was 1.14 (P < 0.001). Compared with women who had never used HRT, those who had used HRT for ≥ 5 years had an increased risk for breast cancer (RR 1.35, 95% CI 1.21 to 1.49, P < 0.001), current users (women who were using HRT at the time of or had used it within 12 months of the diagnosis of breast cancer) had an increased risk (RR 1.21, P < 0.001), and those who had stopped use ≥ 5 years before diagnosis had no increase in risk. Among current users of HRT or those who had ceased use 1 to 4 years previously, the RR increased by 1.02 (CI 1.01 to 1.04, P < 0.001) for each year of use. The RR associated with long durations of current or recent use decreased with increasing weight (P for trend = 0.004) and increasing body mass index (P for trend < 0.001).


Risk for breast cancer is increased in women who are current or recent users of hormone replacement therapy compared with those who have never used it. This risk increases with increasing duration of use. Women who stop using HRT have a reduced risk and have no increase in risk ≥ 5 years after stopping.

Source of funding: Imperial Cancer Research Fund.

For article reprint: Professor V. Beral, Secretariat, ICRF Cancer Epidemiology Unit, Gibson Building, Radcliffe Infirmary, Oxford 0X2 6HE, England, UK. FAX 44-1865-310545.


This detailed reanalysis of epidemiologic study data in breast cancer reminds us that decision making in the use of HRT is complex and that rigorous controlled trials are needed—the Women's Health Initiative Study Group trial is scheduled to be completed in 2007. The main difficulty is that in the studies reviewed, only 12% of the women who used HRT were exposed to progestogens. Thus, the results apply mainly to estrogen alone and not to combination HRT. Other confounding factors include selection of women to receive HRT and screening for breast cancer. High bone-mineral density is associated with increased risk for breast cancer, presumably mediated by lifetime exposure to estrogen. Thus, selection of women with osteoporosis to receive HRT may underestimate the risk for breast cancer associated with HRT because these women have a low risk for breast cancer. Differential selection for screening may also introduce bias. Earlier detection of breast cancer in women who were screened may overestimate the risk for breast cancer associated with HRT. The finding in the review of increased risk for localized but not metastatic cancer is consistent with screening and other effects.

Nevertheless, this review provides the best current evidence that HRT is associated with risk for breast cancer. The risks seem modest: HRT for 5 years is associated with 2 extra cases of breast cancer by age 70 in 1000 women. Cases of breast cancer per 1000 women for 10 and 15 years of exposure are 6 and 12, respectively. These estimates must be put in the context of beneficial and harmful effects of HRT on other organs.

Chris J. Williams, DM
Institute of Health SciencesOxford, England, UK

Chris J. Williams, DM
Institute of Health Sciences
Oxford, England, UK