Background Prospective cohort studies of contemporary populations in both Western and Asian settings have reported a U-shaped association between fertility and mortality. We examined whether an association exists between fertility and all-cause and cause-specific mortality in a sample of Japanese women.

Methods A prospective cohort study was conducted in one rural and one urban community in Gunma Prefecture, Japan, in 1993. A follow-up survey was conducted in the year 2000 in 4858 women aged 47–77 years, and information on demographic and lifestyle characteristics was collected. Mortality and migration data through December 2008 were obtained. A Cox proportional hazard model was used to examine the relationship between parity and mortality.

Results Compared with women with 2–4 children (reference group), higher total mortality was observed among women with 0–1 or 5+ children. Low (HR 1.7, 95% CI 1.2 to 2.3) and high (HR 2.1, 95% CI 1.0 to 4.7) parities were both associated with higher all-cause mortality even after adjusting for potential confounders. Mortality due to ischaemic heart disease exhibited a significant association with parity; the HRs were 3.2 (95% CI 1.1 to 9.2) for women with 0–1 child and 8.7 (95% CI 1.7 to 45.5) for women with 5 or more children. No significant association was observed for mortality from malignancies, cancer of the digestive system, cardiovascular disease or cerebrovascular disease.

Conclusions There exists a U-shaped association between parity and all-cause mortality and cause-specific mortality due to ischaemic heart disease among Japanese women.