Combined oral contraceptives, smoking, and cardiovascular risk.

STUDY OBJECTIVE: To assess age specific incidence and mortality of stroke, acute myocardial infarction (AMI), and idiopathic venous thromboembolism (VTE) associated with use of modern low dose combined oral contraceptives (OCs) and the interaction with smoking. DESIGN: Hospital-based case-control study. SETTING: Hospitals in Oxford region in the United Kingdom, which covered a defined population, during the period 1989-1993. METHODS: Relative risk estimates from the WHO Collaborative Study and observed incidence rates from the Oxford region were used to estimate age specific incidence of each disease among women without cardiovascular risk factors and model total cardiovascular incidence and mortality. RESULTS: Among women who did not use OCs, smoke nor had any other cardiovascular risk factors, total incidence of stroke and AMI were less than 2 events per 100,000 woman years in those aged 20-24 years and rose exponentially with age to 8 events per 100,000 among women aged 40-44 years. Incidence of idiopathic VTE among women who did not use OCs rose linearly with age (from 3.3 per 100,000 at ages 20-24 years to 5.8 per 100,000 at ages 40-44 years). The increased risk of idiopathic VTE associated with OC use among non-smokers constituted over 90% of all cardiovascular events for women aged 20-24 years and more than 60% in those aged 40-44 years. Fatal cardiovascular events were dominated by haemorrhagic stroke and AMI, and among OC users who smoked these two diseases accounted for 80% of cardiovascular deaths among women aged 20-24 years, rising to 97% among those aged 40-44 years. Cardiovascular mortality associated with smoking was greater than that associated with OC use at all ages. Attributable risk associated with OC use was 1 death per 370,000 users annually among women aged 20-24 years, 1 per 170,000 at ages 30-34 years, and 1 per 37,000 at ages 40-44 years. Among smokers, the cardiovascular mortality attributable to OC use was estimated to be about 1 per 100,000 users annually among women aged less than 35 years, and about 1 per 10,000 users annually among those above the age of 35 years. CONCLUSION: The incidence of fatal cardiovascular events among women aged less than 35 years is low. The VTE risk associated with OC use is the largest contributor to OC induced adverse effects. The potentially avoidable excess VTE risk associated with the newer progestogens desogestrel and gestodene would account for a substantial proportion of total cardiovascular morbidity in this age group. For women over age 35 years the absolute risks associated with OC use and smoking are greater because of the steeply rising incidence of arterial diseases. The combination of smoking and OC use among such women is associated with particularly increased risks. Any potential reduction in AMI or stroke risk with use of third generation OCs would be a more important consideration among older compared with younger women, particularly if they smoke. However, the mortality associated with smoking is far greater than that associated with OC use (of any type) at all ages.