Demography, Vol, 29, No.2, May 1992 Birth Spacing and Child Mortality in Bangladesh and the Philippines* Jane E. Miller James Trussell Anne R. Pebley Barbara Vaughan Office of Population Research Princeton University 21 Prospect Avenue Princeton, NJ 08544 This analysis uses data from Bangladesh and the Philippines to demonstrate that children who are born within 15 months of a preceding birth are 60 to 80% more likely than other children to die in the first two years of life, once the confounding effects of prematurity are removed. The risks associated with short conception intervals are confined to children who are also high birth order; they persist in the presence of controls for prior familial child mortality, breast-feeding, mother's age, and socioeconomic status. In Bangladesh but not in the Philippines, these effects are confined to the neonatal period. In recent years a great deal of attention has been focused on whether family planning can significantly improve children's health by shifting births away from the highest-risk maternal age, birth spacing, and parity groups (Bongaarts 1987; Hobcraft 1987; National Research Council 1989; Trussell 1988). Discovering the causal relation between length of preceding birth interval and infant and child survival is complicated by a variety of factors, including confounding of observed bivariate associations by prematurity (DaVanzo, Butz, and Habicht 1983; Miller 1989; Pebley and Millman 1986), breast-feeding patterns (Palloni and Tienda 1986; Trussell and Pebley 1984), prior child mortality (Cleland and Sathar 1984; DeSweemer 1984; Hobcraft, McDonald, and Rutstein 1985; Majumder 1990), poor parenting skills (Das Gupta 1990), and socioeconomic status (National Research Council 1989; Winikoff 1983). The potential for confounding by prematurity has often been downplayed in the literature. Few studies involving data from developing countries have attempted to correct for the problem, largely because of a lack of appropriate data on gestational age. When the * This project was supported by funds from NICHD Grant ROI-HDI7709, and Rockefeller Foundation Grant GA PS 9006. We gratefully acknowledge the International Centre for Diarrhoeal Disease Research, Bangladesh and the Cebu Study Team for providing the data for this analysis. We also thank Andrew Foster for his advice on the use of data from the ICDDRB's Demographic Surveillance System, and Samuel Preston and three anonymous reviewers for their helpful comments. An earlier version of this paper was presented at the 1991 meetings of the Population Association of America in Washington. A more extensive version is available as a working paper (Miller et al. 1991). Copyright © 1992 Population Association of America 305 Downloaded from http://read.dukeupress.edu/demography/article-pdf/29/2/305/904247/305miller.pdf by guest on 29 December 2021