Smoking is associated with increased risk of adverse outcomes in women who use oral contraceptives (OCs), especially in older age groups. New data from continuous monitoring of British women who used OCs since 1968 quantifies this risk and more clearly demarcates women at risk of premature death. Vessey and colleagues report on mortality rates in relation to OC use and smoking.

More than 17,000 women 25 to 39 years of age were recruited between 1968 and 1974 for the study by British family planning clinics. Extensive data were collected from each participant, including reproductive and medical history, demographic information, and health-related habits such as smoking. These data were updated during follow-up visits, and information was added about significant medical events, hospital admissions, and use of hormone therapy. Nonattenders were contacted by mail or telephone to collect data. Contact was maintained at least once per year until each woman reached the age of 45 years.

During the 479,400 woman-years of observation, 889 deaths occurred. Overall mortality was not affected by OC use but was strongly related to smoking. Death rates in heavy smokers were double the rates in nonsmokers and ranged from an excess of 0.7 deaths per 1,000 woman-years in women 35 to 44 years of age to 4.9 per 1,000 in women 55 years or older. Overall the risk of death from ischemic heart disease was slightly but not significantly raised in OC users but was substantially influenced by smoking. Analysis showed no effect of OC use on cardiovascular mortality for nonsmokers. Conversely, in heavy smokers, the rate ratios were 2.4 for OC use of up to 48 months, 4.8 for use of 49 to 96 months, and 2.8 for use of 97 months or longer.

Smoking also was strongly related to mortality from hemorrhagic stroke and increased mortality from other types of stroke and circulatory diseases. Breast cancer mortality did not appear to be related to either OC use or smoking. Cervical cancer deaths were strongly related to OC use, with a rate ratio of 7.2. Cervical cancer deaths also were increased with heavy smoking, but the relationship was not statistically significant. Both ovarian and uterine cancer deaths were significantly reduced in OC users, with rate ratios of 0.4 and 0.2, respectively. No association was found between OC use and melanoma, colorectal cancer, lymphomas, or leukemias.

The authors conclude that the OCs used during the 1970s and early 1980s had no adverse effect on overall mortality in women, but that smoking more than 15 cigarettes per day doubles mortality even in young women. This effect increases with age. The lower hormone content of modern OCs should be associated with even more beneficial effects, but the effects of smoking in women of childbearing age are harmful.