Correlates of non-institutional delivery to delayed initiation of breastfeeding in Nigeria: logit-decomposition and subnational analysis of population-based survey

Abstract

Abstract Background Studies have connected newborn delivery settings and modality to optimal breastfeeding, but how it influences untimely initiation, mostly prevalent in sub-Saharan Africa is unknown. Hence, the role of home delivery on delay initiation of breastfeeding (DIBF) in Nigeria was investigated to inform evidence-based strategy for improved breastfeeding practice. Methodology This is a secondary analysis of births (11,469 home and 7632 facility delivery) by 19,101 reproductive age women in the 2018 NDHS. DIBF is the outcome, home birth is the exposure, and explanatory variables were classified as: socio-demographics, obstetrics and economic factors. Descriptive statistics (frequencies and percentages) were reported, and bivariate (chi-square) analysis was carried out at 20% (p < 0.20) cutoff point. Multivariable logistic regression assessed the probability and significance of the outcome per place of birth. Multivariate decomposition further evaluated the endowment and coefficient effect contribution by independent factors to the outcome. Analysis was carried out at p < 0.05 (95% confidence level) on Stata. Results 56.6% of mothers DIBF, with 37.1% and 19.5% from home and facility delivery, respectively. Home delivery (AOR = 1.34, 95% CI 1.17–1.52) increase the chance of DIBF by 34%, while DIBF probability reduces by 26% in facility delivery (AOR = 0.74, 95% CI 0.65–0.85). DIBF is 5 times more likely in caesarian section delivery (AOR = 5.10, 95% CI 4.08–6.38) compared to virginal birth in facility delivery. Skilled antenatal provider, parity and wealth are negatively associated with DIBF in home birth, while undesired pregnancy, rural residency, partial/no skin-to-skin contact and large child size positively influence DIBF in both home and facility delivery. Skilled antenatal provider (C =  − 66.3%, p < 0.01) and skin-to-skin contact (C =  − 60.6%, p < 0.001) contributed most to reducing the negative DIBF effect with 69% and 31% overall characteristics and coefficient effect component, respectively. DIBF is more likely in Bauchi and Sokoto but less likely in Bayelsa. Conclusions High DIBF prevalent in Nigeria was largely due to elevated rate of home birth, positively associated with DIBF. Caesarian section delivery though heightens the chance of DIBF in facility delivery. Strengthening utilization of skilled provider and skin-to-skin contact can eliminate two-third of the adverse DIBF effect and improve early initiation rate. Adopting this strategy will bridge home-facility delivery gap to achieve optimal breastfeeding practice.