Nielsen, N. M.3; Jorgensen, K. T.3; Stenager, E.4; Jensen, A.3; Pedersen, B. V.3; Hjalgrim, H.3; Kjaer, S. K.3; Frisch, M.3
1 Department of Neurology, Institute of Regional Health Research, Det Sundhedsvidenskabelige Fakultet, SDU2 Center Sønderjylland, Institute of Regional Health Research, Det Sundhedsvidenskabelige Fakultet, SDU3 unknown4 Department of Neurology, Institute of Regional Health Research, Det Sundhedsvidenskabelige Fakultet, SDU
Background: It has been suggested that reproductive factors may be involved in the etiology of multiple sclerosis (MS). We studied associations of reproductive history with MS risk in a population-based setting. Methods: Using national databases, we established a cohort comprising 4.4 million Danish men and women born between 1935 and 1989 and alive in 1968 or later. We obtained information about their live-born children, pregnancy losses, pregnancy complications, and infertility diagnoses. MS cases in the cohort were identified through 2004 in the Danish Register of Multiple Sclerosis. Associations between reproductive factors and MS risk were evaluated using rate ratios (RRs) obtained in log-linear Poisson regression analysis. Results: MS was diagnosed in 6332 women and 3426 men. In both sexes, parents had a lower risk of MS compared with childless persons (in women, RR = 0.76 [95% confidence interval = 0.71-0.82]; in men, 0.89 [0.80-0.98]). RRs were inversely associated with number of children, age at first childbirth, and proximity in time since most recent birth. Among women, MS risk was unrelated to histories of pregnancy loss, pregnancy complications, or infertility. A supplementary analysis in which the date of MS diagnosis was backdated by 5 years to address the possibility of reverse causality did not confirm a protective effect of parenthood (in women, 0.95 [0.88-1.03]; in men, 1.08 [0.98-1.20]). Conclusions: Similar findings in women and men argue against a biologic role of pregnancy in the etiology of MS. Moreover, the observed differences in childbearing patterns were restricted to the 5 years before MS diagnosis, suggesting that reverse causality (ie, reduced reproductive activity in persons with yet-undiagnosed MS) might explain the observed associations.