Objectives: While reproductive factors might plausibly be involved in the etiology of rheumatoid ar-thritis (RA), the female predominance remains unexplained. We aimed to address the possible impact of livebirths, pregnancy losses, and pregnancy complications on subsequent risk of RA in a nationwide cohort study.
Methods: National register data were used to link reproductive histories and later RA hospitalizations in a cohort of 4.4 million Danes. As our measure of relative risk associated with different reproductive histories we used ratios of first inpatient RA hospitalization rates (RRs) with 95% confidence intervals (CIs) obtained in Poisson regression analysis.
Results: Overall, 7,017 women and 3,041 men were hospitalized with RA during 1977-2004 (88.8 mil-lion person-years). RA risk was inversely associated with age at birth of first child in both women and men (P-trend <0.001). Overall, nulliparity and histories of pregnancy loss were not associated with RA risk but, compared with one-child mothers, women with two (RR=0.84; 95% CI: 0.78-0.90) or three (RR=0.83; 0.77-0.91) children were at reduced risk. RA risk was elevated among women with a history of hyperemesis (RR=1.70; 1.06-2.54), gestational hypertension (RR=1.49; 1.06-2.02), or pre-eclampsia (RR=1.42; 1.08-1.84).
Conclusions: One-child mothers and young parents were at increased risk of RA later in life, possibly due to socioeconomic factors. The novel finding of significantly elevated RA risk in women whose pregnancies were complicated by hyperemesis, gestational hypertension, or pre-eclampsia might reflect reduced immune adaptability to pregnancy in RA disposed women or a role of fetal microchimerism in the etiology of RA.