The reduction of regional and socioeconomic inequalities in mortality within countries is a major objective of national governments and international organizations [1–3]. To achieve this goal, determinants of high mortality among disadvantaged people, communities, and regions need to be identified. The physical, ecological, and political structure and impoverished socioeconomic milieu in several countries in sub-Saharan Africa account for geographic variations in childhood mortality [4–6]. One such environment is the regional environment. Poor or polluted environments tend to expose children to disease-causing agents, predisposing them to high mortality risks [7]. In Nigeria, marked regional disparities in mortality of children under age 5 have been reported, with higher rates observed in the Northern regions than in the Southern regions [8–10]. However, these were survey reports rather than empirical studies, and the role of explanatory factors was not investigated. Regional disparities in health-seeking behavior have been reported regarding child immunizations [11], maternal and child health care utilization [12], differences in the socioeconomic composition [13], communicable diseases [13], childhood nutrition, and malnutrition [14]. Other studies have reported that the Northern regions have higher proportions of home delivery and complications during childbirth [15], younger age of first marriage, younger age at birth of first child, and lower knowledge and use of contraception compared to the Southern regions [16].

Regional disparities in these parameters are associated with factors at the community level that distinguish these regions from each other. The availability of services and social amenities in communities, or the lack thereof, may positively or negatively influence the health of the residents of communities. Some of these factors include differences in community-level development, population density, prevalence of poverty, and availability of maternal and child health care services. These are often interrelated aspects of the regional environment that are important for child health and well-being, and may also be relevant in exacerbating or mitigating inequities in resources and population health outcomes across regions [7]. Regions may either exert more independence and autonomy within nations (in which case they become even more salient for child health), or may exhibit greater centralization (in which case regional factors may not be as significant) [7]. A classic example of attempts to attain the former is the South South (Niger Delta) region in Nigeria [17]. This study contributes to a better understanding of how regional disparities in social services and health-seeking behavior influence inequalities in child survival. The aims of this study are two-fold: 1) to determine whether there is significant variation in the risks of under-5 deaths across regions of Nigeria; and 2) to determine the individual- and community-level factors that explain disparities in the risks of under-5 deaths among regions.

Cross-sectional data from the 2003 Nigeria Demographic and Health Survey (DHS) were used in this study. This is a nationally representative sample collected by face-to-face interviews from 3,725 women aged 15 to 49 years with a stratified two-stage cluster sampling procedure. An extensive report of the survey can be found elsewhere [18]. Birth history data, such as month and year of birth, survivorship status, and current age or age at death if the child had died were collected for each birth. This sample contained a total of 6,029 live-born children, limited to births in the last five years before the survey to ensure that the household variables investigated provided a close enough or accurate picture of the current living conditions of the children within period they were exposed to increased risks of death.

This study showed that under-5 mortality was significantly associated with region of residence, with higher risks of under-5 deaths for children of mothers resident in the South South region after adjusting only for individual-level risk factors and only community-level risk factors, as well as simultaneously adjusting for individual- and community-level risk factors. The risk of under-5 deaths was higher in the North East and North West regions after adjusting for child-level variables; however, these differences were reduced and became insignificant once mother-level and community-level controls were added. These elevated risks in the North East and North West regions were explained in part by differences in mothers' education, which corroborates the finding that mothers in the northern regions within this study have a higher proportion of no education or primary education. Findings in this study indicate that demographic factors such as birth order, socioeconomic factors such as marital status, mothers' education, and community-level factors such as living in communities with a low proportion of mothers that received prenatal care by a doctor are the main predictors of regional under-5 mortality. This may be associated with spatial inequality in social development in the community within regions, which may also be associated with population density, differential levels of regional development, political and religious situations, as well as varying economic resources [32]. These factors reflect the situation in the South South region of Nigeria, which is reported to suffer from deficient social infrastructure and services (schools, roads, electricity, and health services), high unemployment, social deprivation, and endemic conflict, in spite of the region accounting for more than 90% of Nigeria's proven gas and oil reserves and the nation's wealth [33, 34]. Geographically, the region is characterized by extensive mangrove forests and extensive networks of lagoons and swamps affected by environmental degradation from crude oil spillage and pollution. These conditions may be related with the increased risks of under-5 deaths for children in this region.

In addition, cross-level interaction between mother's education and community prenatal care by a doctor (i.e., mothers who had no education and lived in communities with low community prenatal care) was associated with a more than twofold increase in risks of under-5 deaths. Although the specific mechanisms underlying the increased mortality are unknown, plausible explanations may include the fact that mothers with little or no education residing in communities deprived of prenatal care by a doctor are more vulnerable because they are generally poorer and lack the economic means for essential goods and services (health care, medications, and transportation) compared to mothers with higher education and access to prenatal care.

The risks of under-5 deaths were higher for ≥ 5 birth order. The influence of birth order on various types of child outcomes is largely dependent on the social and cultural context and may be associated with social disadvantages within families and communities [35], given that large sibships may be a marker for low socioeconomic status [36] and erosion of nutritional resources [37, 38]. However, the influence of other unmeasured factors such as utilization of family planning services cannot be ruled out. This study also found that lower socioeconomic position (divorce, no education, primary-only education) was associated with increased risks of under-5 deaths, which is in agreement with findings from previous studies indicating that higher socioeconomic position of individuals and populations strongly influences health-seeking behavior and is associated with better health [39, 40].

Furthermore, living in communities with a low percentage of mothers who received prenatal care by a doctor was associated with higher risks of under-5 deaths. This could be explained by lower access to prenatal care directly increasing the chances that mothers in the community would not utilize health care services, such as institutional delivery and immunization for their child [41, 42]. Timely access to prenatal care by a doctor is an important preventive measure against maternal and child health outcomes such difficult or obstructed labor, postnatal bleeding, and child deaths [43, 44]. Community prenatal care by a doctor is also an indication of the quality of care received by the mother and infant during childbirth. This association at the community level is also a reflection of socioeconomic position at the individual level because individual socioeconomic position strongly influences health-seeking behavior by enhancing mothers' perceptions of disease etiology and treatment patterns, resulting in improved health and welfare of their children. Higher socioeconomic position empowers mothers and enhances their decision-making power. Community-level variation remained significant after controlling for individual- and community-level variables, indicating that differences among communities still remain unexplained and need further exploration regarding the community-level determinants of under-5 mortality. This finding is also validated by the result of the cross-level interaction.

The intraclass correlation across communities in all the models and across mothers in Model 1 was significantly different from zero, implying that there are significant differences in the risks of under-5 deaths between communities in all the models and between mothers in Model 1 (with region of residence as the only explanatory variable).

The geopolitical regions of low- and middle-income countries such as Nigeria are an important sphere of influence on child survival. As such, findings show the relevance of adopting a more spatially disaggregated, community-level approach to regional policies, given that regions are usually made up of people and communities sharing similar geographical, political, socioeconomic, and cultural characteristics that either promote or inhibit health-seeking behavior and access to health care facilities.

Several limitations need to be considered when interpreting findings in this study. First, administratively defined boundaries were used as a proxy for neighborhoods or communities in this study. There is an inherent risk of nondifferentially misclassifying individuals into inappropriate administrative boundaries, which may generate information biases and reduce the validity of the analysis. Second, data on household income or expenditure, which are the indicators commonly used to measure wealth, are not routinely collected in DHS surveys. The assets-based wealth index used in this study is only a proxy indicator for individual/household economic status and may not always produce results similar to those obtained from direct assessments of income and expenditure where such data are available or can be reliably collected [45, 46]. The strengths of this study are also worth mentioning and include: i) DHS surveys are nationally representative and enable the generalization of the results across the country; ii) the DHS variables are defined similarly across countries, and results are therefore comparable across countries [47, 48]; and iii) using administrative boundaries permits the comparability of any set of DHS data on the same geographical frame, or of presenting complex data in a simple way, provided there is a good conceptual framework of the studied territory.

The authors declare that they have no competing interests.