Effects of Compulsory and Free Basic Education Policies and COVID-19-related Educational Disruptions on Offspring Mortality in Low-Income and Middle-Income Countries: a Quasi-Experimental and Modelling Study.

BACKGROUND:Compulsory and free basic education policies (hereafter called education policies) are essential for enhancing educational attainment, which has been linked to improved offspring health outcomes. However, existing studies do not provide strong evidence with causal inference on how these policies influence offspring mortality in low-income and middle-income countries (LMICs). Furthermore, although COVID-19 has substantially disrupted education systems in LMICs, its long-term consequences for offspring mortality remain unclear. METHODS:We extracted individual-level data from the nationally representative Demographic and Health Surveys from 1986 to 2022, and macro-level data on countries' population projection, COVID-19-related changes, and other social and economic indicators from multiple databases. Offspring deaths were measured using four indicators: neonatal mortality, infant mortality, childhood mortality, and deaths in children younger than 5 years (under-5). We identified 20 countries with at least two survey rounds and changes in education policies during varying survey periods. Using a difference-in-differences design with a two-way fixed-effects model, we investigated the effects of education policies on offspring mortality for mothers and fathers. The treatment group included observations from post-policy surveys, with younger parents exposed to the policy and older ones not, whereas the control group used pre-policy survey data. Subgroup analyses by household wealth status, place of residence, sex of the child, and sex of the household head were performed to explore potential heterogeneity in education policies' impact. We used mathematical modelling to project offspring deaths from 2030 to 2050 attributable to COVID-19-induced loss in educational instruction time and increased school dropout levels, assuming no further policy interventions to mitigate these losses. FINDINGS:A total of 347 013 mothers (100 113 control; 246 900 treatment) and 201 612 fathers (86 152 control; 115 460 treatment) were included in the study. Exposure to education policies was associated with significant reductions in offspring mortality among mothers, which was -11·71 per 1000 livebirths (95% CI -20·29 to -3·12, p=0·0093) for neonatal deaths, -18·66 (-27·06 to -10·25, p<0·0001) for infant deaths, -20·33 (-28·85 to -11·81, p<0·0001) for childhood deaths, and -20·43 (-28·96 to -11·90, p<0·0001) for under-5 deaths, compared with those not exposed to education policies. The effects among fathers were also statistically significant for all types of offspring deaths. Subgroup analyses revealed that education policies have significantly more pronounced impacts among those with lower household wealth. Projections for 2030 to 2050 suggest that reductions in parental education due to COVID-19 might contribute to increases in offspring deaths under a worst-case scenario with no compensatory policy interventions, with the largest effects arising from the loss of parental instruction time rather than from school dropouts. The impact is expected to peak around 2038, when under-5 deaths attributable to reduced parental instruction time and increased school dropouts are projected to reach 449 154 (95% CI 240 774-657 535) and 10 918 (7697-14 139), respectively, compared with a scenario without COVID-19-related educational disruptions. INTERPRETATION:Education policies significantly reduced offspring mortality in LMICs, with greater benefits for populations with lower household wealth. Modelling indicated that interruptions to educational instruction time led to a substantial increase in offspring mortality. There is a critical need for policies to mitigate educational disruptions to prevent offspring deaths. FUNDING:Research Fund, Vanke School of Public Health, Tsinghua University.