Heterogeneous Effects of Birth Spacing on Neonatal Mortality Risks in Bangladesh

Abstract

The negative relationship between birth interval length and neonatal mortality risks is well documented, but heterogeneity in this relationship has been largely ignored. Using the Bangladesh Maternal Mortality and Health Care Survey 2010, this study investigates how the effect of birth interval length on neonatal mortality risks varies by maternal age at birth and maternal education. There is significant variation in the effect of interval length on neonatal mortality along these dimensions. Young mothers and those with little education, both of which make up a large share of the Bangladeshi population, can disproportionately benefit from longer intervals. Because these results were obtained from within‐family models, they are not due to unobservable heterogeneity between mothers. Targeting women with these characteristics may lead to significant improvements in neonatal mortality rates, but there are significant challenges in reaching them.

When and Where Birth Spacing Matters for Child Survival: An International Comparison Using the DHS

“…Nevertheless, other recent research on infant and child mortality using the same statistical approach has found quite different results. Two studies of poor, high-mortality populations—specifically, nineteenth century Stockholm, Sweden, and contemporary Bangladesh—have shown that shorter birth intervals increased the risk of neonatal, postneonatal, and child mortality (Molitoris 2017, 2018b). Furthermore, the latter two studies presented results that may explain the discrepancy in findings mentioned earlier.…”

“…First, the effects of birth interval length on mortality risks decreased over time as the overall level of mortality declined in Sweden (Molitoris 2017). Second, even within a high-mortality context, the size of the effects of interval length on mortality varied inversely with the educational level of the mother in Bangladesh (Molitoris 2018b).…”

“…These research findings call into question whether birth intervals really matter for perinatal outcomes at all (Klebanoff 2017). At the same time, recent research on low-income populations has shown that even after unobserved maternal heterogeneity is adjusted for, birth intervals are still highly consequential for infant mortality in high-mortality populations, such as Bangladesh and sub-Saharan Africa (Kozuki and Walker 2013; Kravdal 2018; Molitoris 2018b).…”

Taking birth year into account when analysing effects of maternal age on child health and other outcomes: The value of a multilevel-multiprocess model compared to a sibling model

“…Another is the study by Bjørngaard et al (2015), who add a crude control for birth year when estimating the importance of maternal age for offspring suicide risks. Kudamatsu (2012) and Molitoris (2018) include both maternal age and birth year in their studies of infant mortality, but without paying much attention to the effects of these two variables, as their focus is on other mortality determinants. However, it is not possible to separate the effects of maternal age and birth year in a sibling analysis without making some bold assumptions.…”

Are sibling models a suitable tool in analyses of how reproductive factors affect child mortality?

“…Many researchers have therefore estimated sibling models, and thus controlled for unobserved parental and environmental characteristics that are shared by the siblings, although it is a disadvantage that one-child families and families where all children have had the same outcomes do not contribute in the estimation. For example, sibling models have been used in several recent studies of the importance of birth intervals for the chance of preterm birth or low birth weight (Ball et al 2014;Shachar et al 2016;Hanley et al 2017;Class et al 2017;Regan et al 2018), child mortality (Molitoris 2018;Molitoris, Barclay, and Kolk 2019), or socioeconomic outcomes (Barclay and Kolk 2017).…”