Objective: To see whether the use of oral contraceptives influences mortality.
Design: Non-randomised cohort study of 17,032 women followed up on an annual basis for an average of nearly 16 years.
Setting: 17 Family planning clinics in England and Scotland.
Subjects: Women recruited during 1968-74. At the time of recruitment each woman was aged 25-39, married, a white British subject, willing to participate, and either a current user of oral contraceptives or a current user of a diaphragm or intrauterine device (without previous exposure to the pill).
Main outcome measures: Overall mortality and cause specific mortality.
Results: 238 Deaths occurred during the follow up period. The main analyses concerned women entering the study while using either oral contraceptives or a diaphragm or intrauterine device. The overall relative risk of death in the oral contraceptive users was 0.9 (95% confidence interval 0.7 to 1.2). Though the numbers of deaths were small in most individual disease categories, the trends observed were generally consistent with findings in other reports. Thus the relative risk of death in the oral contraceptive users was 4.9 (95% confidence interval 0.7 to 230) for cancer of the cervix, 3.3 (95% confidence interval 0.9 to 17.9) for ischaemic heart disease, and 0.4 (95% confidence interval 0.1 to 1.2) for ovarian cancer. There was a linear trend in the death rates from cervical cancer and ovarian cancer (in opposite directions) with total duration of oral contraceptive use. Death rates from breast cancer (relative risk 0.9; 95% confidence interval 0.5 to 1.4) and suicide and probable suicide (relative risk 1.1; 95% confidence interval 0.3 to 3.6) were much the same in the two contraceptive groups. In 1981 the relative risk of death in oral contraceptive users from circulatory diseases as a group was reported to be 4.2 (95% confidence interval 2.3 to 7.7) in the Royal College of General Practitioners oral contraception study. The corresponding relative risk in this study was only 1.5 (95% confidence interval 0.7 to 3.0).
Conclusions: These findings contain no significant evidence of any overall effect of oral contraceptive use on mortality. None the less, only small numbers of deaths occurred during the study period and a significant adverse (or beneficial) overall effect might emerge in the future. Interestingly, the mortality from circulatory disease associated with oral contraceptive use was substantially less than that found in the Royal College of General Practitioners study.
PIP: The objective of this study is to see whether the use of oral contraceptives (OC) influences mortality. A non-randomized cohort study of 17,032 women was followed up on an annual basis for an average of nearly 16 years in 17 family planning clinics in England and Scotland. Women were recruited during 1968-74. At the time of recruitment each woman was aged 25-39, married, a white British subject, willing to participate, and either a current user of OC or a current user of a diaphragm or intrauterine device (without previous exposure to the pill). Overall mortality and cause specific mortality were measured. 238 deaths occurred during the follow-up period. The main analyses concerned women entering the study while using either OC or a diaphragm or intrauterine device. The overall relative risk of death in the OC users was 0.9 (95% confidence interval 0.7 to 1.2). Though the numbers of deaths were small in most individual disease categories, the trends observed were generally consistent with findings in other reports. Thus the relative risk of death in the OC users was 4.9 (95% confidence interval 0.7 to 230) for cancer of the cervix, 3.3 (95% confidence interval 0.9 to 17.9) for ischemic heart disease, and 0.4 (95% confidence interval 0.1 to 1.2) for ovarian cancer. There was a linear trend in the death rates from cervical cancer and ovarian cancer (in opposite directions) with total duration of OC use. Death rates from breast cancer (relative risk 0.9; 95% confidence interval 0.5 to 1.4) and suicide and probable suicide (relative risk 1.1; 95% confidence interval 0.3 to 3.6) were much the same in the 2 contraceptive groups. In 1981 the relative risk of death in OC users from circulatory diseases as a group was reported to be 4.2 (95% confidence interval 2.3 to 7.7) in the Royal College of General Practitioners OC study. The corresponding relative risk in this study was only 1.5 (95% confidence interval 0.7 to 3.0). These findings contain no significant evidence of any overall effect of OC use on mortality. Nonetheless, only small numbers of death occurred during the study period and a significant adverse (or beneficial) overall effect might emerge in the future. Interestingly, the mortality from circulatory disease associated with OC use was substantially that found in the Royal College study. (author's modified).