Influence of Oral Contraception on the Occurrence of Rheumatoid Arthritis in Female Sibs

A matched case-controlled study of the link between oral contraception (OC) and rheumatoid arthritis (RA) was undertaken using siblings as controls, to enhance the similarity of genetic and prior environmental background. This study employed 2 subgroups, a series of affected women from known multicase RA families (119 sisters from 43 sibships from Netherlands), and another series of sporadic RA cases (66 cases and 101 sisters from Britain.) OC use at any time before age of onset of RA in the affected women was compared with OC use at any time before the same age in their sisters. A conditional logistic regression model was used, allowing for multiple cases per stratum; the Mantel-Haenszel odds ratio and the Miettinen's test-based confidence limits were calculated. Confounders considered were marital status, parity, age at menarche, menopausal status, hysterectomy and birth date. 86 cases and 118 controls from 79 sibships were included. 49 sibships were concordant in never using OCs; 8 were concordant in both using OCs. Of the 22 discordant sibships, 6 cases had used OCs and the control sister(s) had not, while in 16 cases the control had used OCs and the index cases had not. The odds ratios calculated with the Mantel-Haenszel method was 0.37, and with conditional regression 0.41, borderline significant (p0.06). The odds ratios were 0.27 combining multicase families, but 0.55 for sporadic case families. These findings are consistent with previously reported studies indicating a protective effect of OCs on the development of rheumatoid arthritis.