Printed in Great Britain Birth Spacing and Child Survival in

Background. Studies examining the associations between short birth spacing and child mortality have often concentrated on the strength of the associations whilst the public health importance of short spacing in specific communities has received less attention. This study re-examines the association between short birth intervals and child mortality in rural Senegal and discusses the potential direct effects of efforts to delay births on child mortality in this community. Methods. The study uses longitudinal data in a cohort of 4852 children born between 1983 and 1989. The associations between birth spacing and child mortality are examined using logistic and Cox proportional hazards regression models. Results. The probability of dying before age five is 224 per 1000 livebirths. The median interval between births is 33 months and only 12 % of the birth intervals are less than 24 months in length. The odds of dying in the neonatal and post-neonatal period is 2.27 and 2.12 times higher respectively for children born after preceding birth intervals of one year or less compared to children born after longer intervals. Children born within two years of a subsequent birth are at 4.09 times higher risk of dying in the second year of life than children whose mother gave birth more than 2 years after the index birth. Conclusions. In this community where prolonged breastfeeding causes women to space their births at long intervals, short birth intervals are a consequence rather than a cause of child mortality and the potential direct effects of birth spacing efforts on child mortality are limited. To reduce the high levels of child mortality, efforts will have to be made to ensure effective preventive and curative health services, and to maintain the traditional pattern of breastfeeding