Determinants of Under-Five Mortality in North East Nigeria: A Multilevel Analysis of Demographic and Health Survey Data

Under-five mortality in Nigeria's north-eastern states remains among the highest globally, driven by a complex interplay of individual, household, and community-level factors. This study employs multilevel logistic regression analysis to identify the hierarchical determinants of under-five mortality using a secondary dataset derived from the 2021 Nigeria Demographic and Health Survey (NDHS), restricted to households in the six north-eastern states. A total of 4,218 child-years at risk were analysed after data cleaning and exclusion of incomplete records. Outcome was defined as death before the fifth birthday. Individual-level predictors included birth interval, birth order, sex, weight at birth, and breastfeeding duration. Household-level variables covered maternal education, wealth index, and water source. Community-level factors included conflict exposure, healthcare facility density, and urbanicity. Multilevel modelling revealed significant clustering of mortality at community level (intraclass correlation coefficient of 0.18). Short birth intervals below 24 months (adjusted OR 3.1), low birth weight (adjusted OR 4.6), and community-level conflict exposure (adjusted OR 2.8) were the strongest independent predictors of mortality. Maternal secondary education was protective (adjusted OR 0.52). Random effects confirmed substantial unexplained community-level variance, underscoring the importance of contextual factors. The study contributes analytical rigour to the evidence base for designing targeted child survival interventions in conflict-affected north-eastern Nigeria. Keywords: under-five mortality, multilevel analysis, NDHS, North East Nigeria, determinants.