Recent reports of an increased risk of breast cancer in women taking continuous combined hormone replacement therapy (HRT) caused significant controversy and distress for the millions of American women using hormones postmenopausally. Much more information is needed about the effect of different forms of HRT on breast cancer and other health risks. Weiss and colleagues studied nearly 4,000 postmenopausal women 35 to 64 years of age to determine odds ratios for breast cancer risk with different forms of HRT.

The authors identified case patients as women with newly diagnosed primary invasive breast cancer at five participating medical centers. Extensive interviews were conducted with 4,575 case patients, covering reproductive and medical history, family cancer history, and relevant lifestyle and demographic information. Interviews also were conducted with 4,682 control patients selected to represent the same distribution of race, age, and area of residence. The study reported data on women who had used any form of HRT. Using statistical models based on no use of hormone therapy, and controlling for potentially confounding variables, the researchers calculated odds ratios for breast cancer following use of estrogen alone, sequential HRT (five to 14 days per month of progestin), and continuous combined HRT (25 days or more of progestin per month).

A total of 1,252 case patients and 1,374 control patients reported use of HRT. Black women (one third of the women in both groups), were much less likely to report use of HRT than white women (48 percent versus 75 percent). Case patients were older than control patients and were more likely to report a family history of cancer, natural menopause, and recent screening mammography. Conversely, case patients were less likely to report use of contraceptives or young age at first pregnancy. In other respects, the groups were similar.

Increasing risk for breast cancer with duration of HRT was found only in patients taking continuous combined formulations. As shown in the accompanying table on page 1060, the odds ratio for estrogen alone was 0.83 for less than six months to 0.84 for five years or more, indicating a lower rate of breast cancer in HRT users than in control patients. For all combined HRT, the odds ratios rose from 0.65 for short durations of use to 1.17 for five years or longer, but this risk changed when continuous and sequential regimens were considered separately. For sequential regimens, the odds ratio at five years or more was 0.96, indicating close to or slightly less than the risk of nonusers.

The authors conclude that the risk of breast cancer varies with the HRT regimen used and is not increased for estrogen alone. The only significantly increased risk in this study was with use of continuous combined HRT for five years or more. This result confirms that of one previous study, but many studies on the association between HRT and breast cancer have given conflicting results. The authors state that the absolute risk of breast cancer may be low for individual women and that much more research is needed about the effect of different regimens of HRT.