|A Bayesian semiparametric multilevel survival modelling of age at first birth in Nigeria|
||Ezra Gayawan, and Samson B. Adebayo
||Demographic Research, 28(45): 1339-1372; DOI: 10.4054/DemRes.2013.28.45
The age at which childbearing begins influences the total number of children a woman
bears throughout her reproductive period, in the absence of any active fertility control.
For countries in sub-Saharan Africa where contraceptive prevalence rate is still low,
younger ages at first birth tend to increase the number of children a woman will have
thereby hindering the process of fertility decline. Research has also shown that early
childbearing can endanger the health of the mother and her offspring, which can in turn
lead to high child and maternal mortality.
In this paper, an attempt was made to explore possible trends, geographical variation
and determinants of timing of first birth in Nigeria, using the 1999 – 2008 Nigeria
Demographic and Health Survey data sets.
A structured additive survival model for continuous time data, an approach that
simultaneously estimates the nonlinear effect of metrical covariates, fixed effects,
spatial effects and smoothing parameters within a Bayesian context in one step is
employed for all estimations. All analyses were carried out using BayesX – a software
package for Bayesian modelling techniques.
Results from this paper reveal that variation in age at first birth in Nigeria is
determined more by individual household than by community, and that substantial
geographical variations in timing of first birth also exist.
These findings can guide policymakers in identifying states or districts that are
associated with significant risk of early childbirth, which can in turn be used in
designing effective strategies and in decision making. These findings can also point in
the direction of effective utilisation of scarce resources: a major challenge to those
effecting interventions in developing countries.