Introduction

In 1996 a large survey reported a slightly increased likelihood of breast cancer occurring in women who had used oral contraceptives over the previous 10 years (relative risk [RR] = 1.24, 95% C.I. 1.15 to 1.33). The findings were based on data pooled from 54 epidemiologic studies conducted over 25 years. Since then, the view has persisted that women taking oral contraceptives are at increased risk.

A new interview-based survey has just been published in the New England Journal of Medicine that should set a lot of minds at rest. It comes from the Centers for Disease Control and Prevention, Atlanta, working with regional centers in 5 major US cities.

Methods

Women between 35 and 64 living in the 5 cities who had invasive breast cancer diagnosed between 1994 and 1998 were identified through local registries. Of 5,972 women identified, 4,575 (76%) were interviewed.

To build a control group, random-digit phone dialing in the same residential districts identified women without invasive or in situ breast cancer. Those identified were sampled randomly within strata defined according to the study site, race, and age. Of 5,956 women selected as controls, 4,682 (79%) were interviewed.

The interviews were conducted face-to-face using a standard questionnaire. The date of initial diagnosis of breast cancer (or, for controls, the date of the telephone screening), details of oral contraceptive use, family history, and reproductive history were obtained.

Oral contraceptives were classified as 'combination' if they included estrogen and progestin in each cycle (whatever the sequence), or as 'progestin-only'. Estrogen dosage was termed 'high' if they contained 50 micrograms or more of ethinyl estradiol or its equivalent. In women with exact information, 95.5% of contraceptive use was of the combination type.

Analyses were designed to calculate the relative risk of breast cancer for different classes of subjects (e.g. present oral contraceptive users vs. never used), using regression analysis to adjust for possible confounding variables.

Results

Oral contraceptive use was similar in the two main groups -- 77% in case subjects and 79% in controls. Of the total, 35% were black.

The risk of breast cancer was unchanged in women who had ever used any type of oral contraceptive compare with controls (RR = 0.9, 95% C.I. 0.8 to 1.0). In women who were current users of oral contraceptives, there was also no increased risk (RR = 1.0, 95% C.I. 0.8 to 1.3).

Analyses for different aspects of oral contraceptive use (any use, current use, former use, duration of use, age at first use, interval since last use, estrogen dosage) revealed no relevant evidence that the risk of breast cancer was increased in users.

Age, menopausal state, race, family history of breast cancer, and body mass index were without influence on the overall findings.

Comment

The authors of this study discuss possible reasons for the discrepancy they have found compared with the result of the 1996 analysis of trials available at that time. Although there were several minor differences in data collection that might have influenced the findings, none were sufficiently important to detract from the principal finding, namely that current or former use of oral contraceptives by women aged 35 to 64 did not significantly increase the risk of breast cancer.

It should not be forgotten that oral contraceptive use is associated with a number of rare, but serious side effects. These include deep venous thrombosis, pulmonary embolism, ischemic stroke, liver cancer, and among women over 35 years of age who smoke, myocardial infarction.