Child-Spacing Effects on Infant and Early Child Mortality

Abstract

For at least 60 years the deleterious effects of rapid childbearing on infant and early child survival chances have been the subject of study. The analyses of child spacing effects on infant and early childhood mortality presented here are based upon data from 26 developing countries which participated in the World Fertility Survey (WFS). All WFS data sets include a complete live birth history including the date of birth of all children born alive and information on their survival status and where relevant age at death. This information allows greater flexibility in analysis. Birth intervals can be calculated and counts of births in specific segments of time can be computed. It is also possible to estimate survivorship status of a birth at the time of some other event such as a subsequent birth or an anniversary of that subsequent birth. The goal is to explore some of the effects of child spacing both in terms of overall but localized rapidity and in terms of specific intervals. The approach adopted is to form counts of births in segments of time before and after the child in question or index child. 3 2-year of observation immediately prior to the birth of the index child are used forming counts of births 0-2 2-4 and 4-6 years before. The approach through the use of windows is designed for multivariate analysis. A table shows exponentiated parameter estimates or relative risks which result from fitting a common model to the data for each country. This model includes main effects for each of the 3 variables considered birth 0-2 and 2-6 years before the index child and mothers education. In general all 3 main effects were significant. The exponentiated parameter estimates for births in the 2 years preceding the index birth were startling. 1 such birth raised the risks by 50% or more in all but 2 countries and more than doubled the risk in 14 countries. These effects were consistently around 2 in Africa and over 2 in most of the Asian countries. 2 or more nearby births in the 2 years preceding the index birth had even more marked effects on neonatal mortality raising it at least 3-fold in 17 countries. In regard to the effects of childbearing experience in the 2-6 years before the index birth occurrence of zero births in this period led to varied relative risks compared with the baseline group. This zero group seemed to experience slightly lower mortality risks in most Latin American countries but little else can be said about patterns.