Risk factors and a predictive model for under-five mortality in Nigeria: evidence from Nigeria demographic and health survey

Childhood mortality is a prominent public health issue. Globally, mortality in under-five age children is about 9 million deaths per year and 70% are preventable which makes this an important public health problem to investigate [1]. In one of the series on child survival published on Lancet, childhood mortality was described as public-health disaster [2]. Children in developing countries are 10 times as more likely to die before their fifth birthday compared to those in developed countries [1]. The gap in childhood mortality between developed nations and developing nations has been increasing over a long period of time [3]. In response to the burden of childhood mortality in sub-Saharan Africa countries, United Nations and World Health Organization (WHO) has incorporated reduction of childhood mortality by two-third by the year 2015 as one of its Millennium Development Goals (MDGs).

Identifying determinants of under-five mortality (U5M) is essential for formulating appropriate health programmes and policies in order to meet the United Nations MDG goal (i.e. to reduce childhood mortality by two-third by 2015). Several studies have been conducted to identify the determinants of U5M. Studies have identified strong associations between maternal factors and childhood mortality. Maternal age has been shown to have a strong association with child survival [4‐8]. However some of these studies did not consider certain important covariates such as sanitation, immunization, breastfeeding and previous child deaths in their research which might have confounding effect. Studies have also shown that children of educated mothers will survive better than children of non-educated mothers [5, 9‐15]. However another author has reported that maternal education has little or no effect on child survival [16].

Birth interval has been identified by numerous studies as a major determinant of childhood mortality [4, 7, 9, 13, 17‐20] although not all of these studies have considered potential confounders like breastfeeding, intra-familial mortality risks (survival status of preceding child), and multiple pregnancy etc. Grand multiparity has been revealed to have a negative effect on child survival [10, 15] Similarly, large family size has also been indicated to have a close association with poor rates of child survival. This was explained by increased intra-familiar competition for foods and other limited resources essential for child survival [13, 19]. Availability of accessible health care services have been reported to have an inverse relationship with childhood mortality [4, 10, 21, 22]. Despite these preponderance of evidence about the determinants of childhood mortality from different parts of the world, the majority of the studies did not consider the potential effects of some of the important confounders in their analyses [4‐9, 13, 18, 20]. Moreover, there are few good studies from Nigeria and most of which did not use national representative data [23‐28]. Thus, the generalizability of the results is a serious problem in a country like Nigeria because of her diversity. The most recent national representative 2008 Nigerian Demographic and Health Survey (NDHS) [29] has estimated U5M rate at 157 children per 1,000 live births compared to their counterpart in the United Kingdom (England and Wales) which was said to be 2.8 children per 1,000 live births [30]. It means that 1 out of 6 children will before their fifth birthday in Nigeria [29]. It is therefore important to the explore the latest Nigeria Demographic and Health Survey data to identify the determinants responsible for the persistently high under-5 mortality in Nigeria. The objective of this study was to develop a predictive model and identify maternal, child, family and other risk factors associated U5M in Nigeria.

The outcome of this study has shed more light on risk factors of U5M in Nigeria. This study has attempted to control for the effects of potential confounders by incorporating many factors into the analyses without over-modelling. Maternal age at first marriage was found to have an inverse relationship with U5M. This result shows that children of mothers that started childbearing at an early age were more prone to experience U5M more than those that commenced at an older age. This problem is now more pronounced because younger mothers formerly derived supports from mother in-laws and relatives that will stay with them for a long period of time but due to the economic crisis, this tradition has been eroded so young inexperience mothers have to take care of their children by themselves. Effect of maternal age at first marriage is consistent with other previous studies [18, 44].

The likelihood of U5M reduced in children of mothers that embraced the use of contraception. This study revealed that U5M were less likely to occur in children of mothers that adopted folkloric or traditional method of contraception than those that were not using any contraception. Contraception has a way of militating against incidence of unwanted pregnancy and it enhances adequate child spacing. It ensures that only children that were planned for will come to life and these will improve their survival. This relationship did not contradict what prior studies have reported [8, 13].

Birth interval was another important predictor of child survival. It has negative relationship with U5M. As child spacing increased, the likelihood of U5M reduced. This result supported what prior studies have reported [4, 7]. Possible explanation for this is that mothers that waited for more than 18 months before having the next baby would have regained most body nutrients and blood loss during previous pregnancy and breast feeding. It is also a common medical knowledge that risk of obstetrics complications is higher in mothers that had short birth interval than those with long birth interval. Birth order was another determinant of child survival. It has a parallel association with U5M. The likelihood of U5M increased with increased in birth order of the child. This is consistent with other previous studies [5, 45]. Possible mechanism for this is that as birth order increases, intra-familiar competition for foods and other limited resources essential for child survival will increase. Moreover, children are more prone to receive most impacts of it. Also as birth order increases level of child care reduces since the mother will have more children to care for.

It is important to mention that this study also revealed that children that were small at birth were more likely to die during before the age of 5 years than those that were having normal birth weight. A possible explanation for this, are the effects of preterm births. Most preterm children are more prone to have sepsis which is one of the leading causes of neonatal deaths. Other complications of preterm births are neonatal jaundice, apnoea etc. All these will account for high the likelihood of U5M in children with low birth weight. This is consistent with other previous studies [7, 8, 10]. This study has also shown that children raised in urban areas were more likely to survive better than those in rural areas. This finding supported what have been reported in other studies [4, 9, 46]. Likely explanation for this relationship is that children in urban areas have better access to health care services and all other essential health related services which are important for child survival.

This study revealed that good toilet facilities at home can reduce U5M and this is in line with other prior studies [21, 47, 48]. The outcome of this research has also shown that the number of wives that men married to, has a positive relationship with U5M. It was observed that as the number of wives increased, the likelihood of U5M in that family increased. This is consistent with previous study [49]. Possible reasons for this is intra-familiar competition for food and other resources which are essential for child survival. Family size was noticed to have an independent effect on U5M. This phenomenon can be explained, based on the fact that as family size increases, intra-familiar competition for foods and other essential services require for child survival increases, so also parental attention for their children decreases. As a result of these U5M will increase.

Health seeking of mothers has been shown to have a positive impact on child survival. It is common medical knowledge that children of mothers that were in a good habit of seeking medical attention will allow their children to have the privilege of receiving immunization and treatment of preventable childhood diseases. Studies have shown that these services reduced U5M [4, 10]. However, the trend was not found in this relationship. It is essential to report that this research yielded two skewed results although this could be explained. It was noticed that U5M started to increase from in mothers in the age group 26-30 years and the odds of U5M increased as the age group increased. Prior studies have reported two contrary views about the effect of maternal age on U5M. Some research reported that maternal aged has an inverse relationship with U5M [23, 26] while others revealed that it has no effect on U5M [7, 50]. However, it is crucial to state that none of these studies considered maternal age at first marriage in their analyses, which was adjusted for in this study. Similarly, this study also revealed that child survival in homes that were using kerosene, fire woods and charcoals as sources of fuel would be better than those using gas. This was not in agreement with previous study [51]. Possible explanation for this is that, < 1% of the studied population were using gas.

Moreover survival of children living in South west of Nigeria was better than those residing in North central of Nigeria. Children in North central of Nigeria were more likely to die before their 5th year birthday than their counterparts in North east of Nigeria. Prior study has also reported regional variations in child survival [52]. Possible explanation for these regional differences in child survival can be attributed to variations in vegetation across the regions. Children tend to survive better in regions with better agricultural output than regions with low output. Also the political will of government to implement good health policies varies across the regions.

This study has revealed that maternal, child and family and other factors were the important determinants of U5M in Nigeria. It will be of great benefit if all these predictors could be properly considered during planning, formulation and implementation of policies that will improve child survival.