Abstract
From data available at April 1987 it was found that the standardised risk ratio for rheumatoid arthritis between current users of oral contraceptives and never users was 0.82 (95% confidence interval 0.59 to 1.15); the ratio between former users and never users was 0.94 (95% confidence interval 0.72 to 1.22). Important secular trends have occurred within our study population. The incidence of rheumatoid arthritis among former and never users has declined over the past two decades. Current users have not experienced this temporal trend, and the ratio between current and never users has, therefore, approached unity. These secular changes may explain why some studies have found that oral contraceptives have a protective effect, while others have been unable to show such an effect.