From parental education to under-5 mortality: how antenatal care and hand hygiene mediate the pathway mechanisms

Background: Reducing the under-5 mortality rate (U5MR) to as low as 25/1000 live births globally by 2030 to reach the sustainable development goal (SDG) 3.2.1 appears to be a race against time. Amidst this situation, Bangladesh experienced stagnancy in U5MR at the beginning of the SDG era. While a comprehensive understanding of this stagnancy is crucial, research on stagnancy is scant and limited to investigating the direct effects of the key predictors. Therefore, building on the existing evidence, this research investigated the effect of one of the key predictors, parental education, on under-5 mortality (U5M) mediated through antenatal care (ANC) and the availability of hand washing stations. Methods: This study conducted weighted mediation analyses using the Bangladesh Demographic and Health Survey data from 2011, 2014 and 2017–18, with sample sizes 4,658, 4,367, and 4,826, respectively, and estimated a generalised structural equation model. The “nlcom” post-estimation command of STATA was used to obtain the direct effects (coefficients), which were then multiplied to estimate the indirect effects for the indirect parallel and sequential pathways. The indirect effects were added to obtain the total indirect effect, which was added to the direct effect of education to estimate the total effect of education. Results: This study reveals inconsistent parallel and sequential mediation of parental education’s effect across the three surveys. The direct effect of education on U5M reduced over time. The total (parallel and sequential combined) mediated effects of either parent’s education on U5M were negative (reduced U5M risk). While the mediated effect of mother’s education on U5M was not significant, the mediated effect of father’s education was significant in 2011. The total effect of education (direct and indirect combined) on U5M was negative (reduced U5M risk) and significant in 2011. However, the effect reduced over time and changed direction in 2017-18. Since 2011, the effects decreased and changed direction to become positive (increased U5M risk) in 2017-18. Due to the weakening direct effect of education and the opposite nature of the indirect effect, the indirect effect of education appeared to be greater than the total effect in 2014 and 2017–18. In 2011, approximately 50% of the total effect of either parent’s education was mediated. However, in 2014 and 2017–18, the proportions were greater than 100%, except for the proportion of the mediated effect of father’s education in 2014. Conclusions: This study reports a weakening and inconsistent mediated effect of both parents’ education on U5M. Using existing literature, it also justifies that improving the quality of education and ANC could more effectively reduce U5M to achieve SDG 3.2.1. To improve the quality of education, this study recommends updating the school curriculum with a greater emphasis on maternal and child health education. To improve the quality of ANC, it is recommended that the mandatory use of ANC cards be strictly monitored.