The effect of reproductive behavior on infant and early childhood mortality in Jordan.

Abstract

This study integrates various aspects of earlier studies of child mortality into a comprehensive examination of the reproductive factors affecting infant and early childhood mortality in Jordan. The probability of dying of an infant is studied over a series of short age intervals employing logit models. The major features of this study include: a detailed examination of the quality of data; utilization of individual, household and community data; tracing the risk of dying over age; and estimating infant mortality for different patterns of reproduction. This study employs data from the 1976 Jordan Fertility Survey, in which the units of analysis are live births which took place in the 5-year period 1971-76. The main findings of the study are that the duration of breast-feeding and the length of the preceding birth interval are important determinants of mortality during infancy. Their effects are quite important shortly after birth, but decline gradually with age, until they are unimportant at age 8 months. In general, the over-all effect of the set of biological and demographic variables such as the preceding birth interval, maternal age, sex of the child, etc, is not important during the last few months of infancy and in the second year. Parents' characteristics such as education appear to be important only for early infancy and in the second year of age. The effects of the household characteristics are significant only in a few age intervals, while community characteristics are not important for any of the age intervals studied. Projections of infant mortality utilizing these results indicate that the potential gains due to changes in birth spacing can be substantial, especially if births are spaced by at least two years.