The study examined how socioeconomic, demographic and cultural factors (also referred to as
underlying factors), influence under-five mortality by identifying some other more direct factors
(called proximate determinants) through which they operate in order to influence under-five
mortality in Nigeria, The data used were from the Nigeria Demographic and Health Survey
(NDHS) of 2013. It was a nationally representative sample. Mothers of the 31,482 children who
were born during the previous five years of the survey were interviewed on the survival status of
their children and also on their own characteristics and child healthcare practices. The conceptual
framework adopted to explore the pathways of influence of the underlying factors on under-five
mortality was developed by Mosley and Chen (1984) in their study of the determinants of child
survival. The statistical analyses used were the univariate, bivariate and the multivariate analyses.
Due to the binary nature of the dependent variable, the multivariate analysis technique adopted
was the “binary logistic regression”. The level of significance was set at 5 percent. Findings
showed that all the nine underlying factors considered in the study had significant indirect bivariate
relationship with the outcome variable. The underlying factors are mother’s age, marital status,
mother’s education, her husband’s education, her occupation, her geographical region, her
urban/rural residence, her wealth status and religious affiliation. At the multivariate analysis level,
all the underlying factors except religion and occupation, maintained their significant indirect
effects on under-five mortality. The seven proximate determinants examined in the study are, the
child’s birth order, number of births the mother had within the previous five years of the survey,
whether or not the child was ever breastfed, the household source of drinking water, household
type of toilet facility, household type of cooking fuel, and the place where the child was delivered
(i.e., in a modern healthcare facility or elsewhere). Findings showed that of the 31,482 under-five
children reported born during the period, 2886 had died giving a mortality rate of 92 per 1000. All
the nine underlying factors indicated significant direct bivariate relationship with six of the seven
proximate determinants, implying that the underlying factors may be influencing under-five
mortality through those proximate variables. The only proximate determinant that many of the
underlying factors did not indicate significant bivariate relationship with was the “ever breastfed”
variable. Only four underlying factors had significant bivariate relationship with it. The four
variables are mother’s education, marital status, religion and geographical region. All the other
five underlying factors did not influence under-five mortality through ever breastfed variable. At
the multivariate analysis level, we found that children who were never breastfed were twenty times
more likely to die before five years of age than children who were breastfed (Chi-square = 21.9 on
1 degree of freedom, P = 0.000). The Log Likelihood Ratio test (LLR) carried out to examine the
overall level of importance of the proximate determinants as conduits or pathways of influence for
the underlying factors on under-five mortality found them to be highly significant (Chi-square =
3075 on 12 degrees of freedom, P = 0.000). However, several of the underlying factors still
maintained their significant relationship with under-five mortality even with adjustments for the
seven proximate variables. The result implies that, in addition to the seven proximate determinants
used in the study, there could be other proximate variables not included in the study through which
the underlying factors may also be influencing under-five mortality. Policies and programs to
reduce under-five mortality which are targeted to influence proximate determinants (e.g., birth
spacing), are cheaper and quicker to produce desired results than are policies and programs that
are aimed at changing the underlying factors, such as increasing levels of women’s education. One
of the recommendations from the findings of this study is that policies and programs be targeted
towards ensuring proper spacing of births by women, particularly those who are 20-39 years old,
3
to prevent too many births taking place within a short period of time. Programs that create
awareness about some of the causes of childhood mortality and also encourage the use of
contraceptives and extensive breastfeeding to prevent frequent or multiple births within short
intervals of time will go a long way in helping to reduce the high under-five mortality in Nigeria.