Skip to main content
NCBI home page
Maternal & Child Nutrition logoLink to Maternal & Child Nutrition
. 2018 Nov 8;15(2):e12722. doi: 10.1111/mcn.12722

Nutritional status as a central determinant of child mortality in sub‐Saharan Africa: A quantitative conceptual framework

Cristian Ricci 1,✉, Janet Carboo 1,*, Hannah Asare 1,*, Cornelius M Smuts 1, Robin Dolman, Martani Lombard
PMCID: PMC7199068  PMID: 30316202

Abstract

Child mortality is a major public health problem in sub‐Saharan Africa and is influenced by nutritional status. A conceptual framework was proposed to explain factors related to undernutrition. Previously proposed conceptual frameworks for undernutrition do not consider child mortality and describe factors related to undernutrition from a qualitative viewpoint only. A structural equation modelling approach was applied to the data from World Bank and FAO databases collected from over 37 sub‐Saharan countries from 2000 to the most recent update. Ten food groups, exclusive breastfeeding, poverty and illiteracy rates, and environmental hygiene were investigated in relation to underweight, stunting, low birthweight, and child mortality. Standardized beta coefficient was reported, and graphical models were used to depict the relations among factors related to under‐five mortality in sub‐Saharan Africa. Child mortality in sub‐Saharan Africa ranged between 76 and 127 × 1,000. In the same period, low birthweight rate was about 14%. Poverty and illiteracy are confirmed to affect health resources, which in turn influenced nutritional status and child mortality. Among nutritional factors, exclusive breastfeeding had a greater influence than food availability. Low birthweight, more than underweight and stunting, influenced child mortality.

Structural equation modelling is a suitable way to disentangle the complex quantitative framework among factors determining child mortality in sub‐Saharan Africa. Acting on poverty at the base appear to be the more effective strategy along with improvement of breastfeeding practice and improvement of hygiene conditions.

Keywords: child mortality, child nutritional status, conceptual framework, sub‐Saharan Africa

Key messages.

  • Child nutritional status is significantly related to child mortality in sub‐Saharan Africa. Low birthweight appeared to be one of its most important determinants.

  • Breastfeeding, more than food availability, determines child nutritional status in sub‐Saharan Africa showing the importance of nutrition in early phases of life.

  • Health resources and environmental hygiene plays an important role in determining child nutritional status in sub‐Saharan Africa.

  • Poverty can be confirmed to act as a basic determinant of child nutritional status and in turn child mortality in sub‐Saharan Africa.

1. INTRODUCTION

Child mortality in sub‐Saharan Africa is a current problem (Jamison, 2006). It was recently reported that up to 50% of the total under‐five mortality occurred in sub‐Saharan Africa in 2013; this corresponded to more than 3 million deaths. During 2013, more than 8,000 children under the age of 5 years died every day in sub‐Saharan Africa (Liu et al., 2015; You, Hug, Ejdemyr, & Beise, 2015). None of this appears to augur well for the future of the children in sub‐Saharan Africa where projections to 2030 estimates that sub‐Saharan Africa will contribute with up to 60% to the overall mortality of children under the age of 5 years (Liu et al., 2015). Under‐five mortality in sub‐Saharan Africa could be attributed to many different factors (Black, Morris, & Bryce, 2003; Foster & Williamson, 2000; Jamison, 2006; Mosley, 1983). Among the others, child nutritional status is of particular interest representing a modifiable risk factor that can be related to up to 20% of child mortality occurring in sub‐Saharan Africa (Lartey, 2008; Lim et al., 2012; Pelletier, 1994). Conceptual frameworks about the determinants of undernutrition in children has been proposed in the past. The most commonly accepted one was the conceptual framework proposed by Engle, Menon, and Haddad (1999), which was also adopted by the UNICEF and other institutions. According to this conceptual framework, three different levels of undernutrition determinants can be defined. Poverty at the base is defined as the basic determinant of undernutrition affecting food availability and security, resources for health care of the mother and the child, and the hygiene of the environment. These latter are defined as underlying determinants of undernutrition and in turn affect child health and child nutritional status. Finally, child health and child nutritional status are defined as immediate determinants of undernutrition. This conceptual framework well represents the factors that determine undernutrition according to a hierarchical structure between them. Nevertheless, it has certain limitations. First, this framework does not include under‐five mortality, which should be placed on the top to complete the framework. Moreover, if child mortality is included at the top of the framework, all the factors included in the conceptual framework should refer to under‐five mortality as a final stage. Second, all of the factors at a given level (basic, underlying, and immediate determinants) appear to have the same importance, although it would be interesting to understand their reciprocal relations. In the present work, we proposed a modelling approach aimed to disentangle the quantitative complex framework among those determinants of child mortality in a revised conceptual framework having similar elements than the one early proposed by Engle, Menon, and Haddad. To this aim, data from the World Bank and FAO databases were collected and analysed using structural equation models.

2. METHODS

The most recent up‐to‐date information about under‐five child mortality and its determinants were collected from the World Bank and FAO databases during the period 2000–2016 and 2000–2013, respectively. A panel of 24 variables was used. Briefly, under‐five mortality, defined as the probability per 1,000 that a newborn baby will die before reaching age 5, was the main outcome. The underweight prevalence, wasting prevalence (percentage of children under age 5 whose weight for age and height for age are more than two standard deviations below the median for the international reference population ages 0–59 months), and low birthweight rate (newborns weighing less than 2,500 g) were considered as immediate determinants of child mortality. Afterwards, health expenditure, hospital beds, and a number of nurses, midwives, and physicians were considered to portray the availability of care resources. On the other hand, the distribution of hand washing and safe water drinking facilities and number of people practicing open defecation were investigated to represent environmental hygiene and sanitation. Percentage of exclusive breastfeeding from the World Bank database and nine groups of foods from the FAO database (animal fat, fish, eggs, meat, milk, vegetable oils, fruit, cereals, and vegetables) were considered to represent food availability. Finally, the gross national income per capita, the unemployment and literacy rates, and national expenditure on education were defined as basic determinants of child mortality. Variables and their definitions from the corresponding databases were reported on Table S1. Thirty‐seven countries from sub‐Saharan Africa had a satisfactory data panel for all variables and could be included in the analysis. Briefly, data were available for Djibouti, Ethiopia, Kenya, Madagascar, Malawi, Mozambique, Rwanda, and Uganda in Eastern Africa. Western Africa was represented by Benin, Burkina Faso, Cote d'Ivoire, The Gambia, Ghana, Guinea, Liberia, Mali, Mauritania, Niger, Nigeria, Senegal, Sierra Leone, South Sudan, and Togo. Central Africa was represented by Angola, Cameroon, Central African Republic, Chad, Democratic Republic of Congo, Gabon, and Sao Tome and Principe. Finally, Botswana, Lesotho, Namibia, South Africa, Swaziland, Zambia, and Zimbabwe were included to represent Southern Africa.

2.1. Statistical methods

Medians and interquartile ranges for under 5‐year mortality and all of the variables considered were reported for both sub‐Saharan African areas and the total sub‐Saharan Africa (Table 1). Before the analysis, all of the variables considered were summarized in a single, time detrended, value for each country. Briefly, the time series were interpolated using a random intercept model applied to each country. The resulting predicted values were then summarized by country using a single value performed by their median. Finally, these values were normalized and standardized using the Blom transformation (Blom, 1958). This step was performed to the aim of having standardized beta regression coefficients to depict the path of association between an outcome and a predictor after structural equation modelling (SEM). Notably, these coefficients can then be more easily interpreted as explained variance having a standardized range of variability between −1 and 1. Finally, after SEM analyses, the absolute value of the point estimate of the standardized beta coefficients was reported to depict the strength of the association between a predictor and an outcome. We hereby considered a value of the module of a standardized beta coefficient between 0.3 and 0.6 as representative of a moderate association between variables. Standardized beta coefficients higher than 0.6 could be interpreted as a strong relation between the outcome and the predictor.

Table 1.

Median and interquartile range over the observational period of the variables considered by areas of sub‐Saharan Africa

Eastern Africa Western Africa Central Africa Southern Africa Sub‐Saharan Africa
Under 5 mortality × 1,000 76.4 (62.8, 113) 110 (86.5, 146) 127 (102, 158) 103 (92.9, 118) 106 (82.0, 138)
Immediate determinants of child mortality
Underweight prevalence (%) 18.4 (16.4, 25.2) 21.1 (16.5, 27.4) 20.5 (15.0, 28.0) 11.6 (9.6, 14.9) 18.4 (14.6, 25.4)
Stunting prevalence (%) 44.3 (39.1, 49.5) 35.6 (27.9, 40.0) 38.7 (31.7, 44.0) 32.8 (29.5, 39.5) 36.6 (29.6, 43.6)
Low birthweight rate × 1,000 16.0 (14.0, 18.1) 14.1 (11.7, 16.2) 13.4 (12.0, 16.7) 11.4 (9.5, 14.2) 13.9 (11.2, 16.4)
Underlying determinants of child mortality
Health expenditure (% GDP) 5.5 (4.6, 7.6) 5.1 (3.9, 6.9) 4.5 (3.5, 5.2) 5.9 (4.6, 7.2) 5.2 (4.2, 6.9)
Hospital beds × 1,000 0.8 (0.6, 0.9) 1.1 (0.5, 1.4) 0.9 (0.4, 1.3) 1.7 (0.7, 2.0) 0.9 (0.5, 1.6)
Nurses and midwives × 1,000 0.9 (0.4, 4.0) 0.5 (0.3, 0.7) 0.5 (0.4, 0.9) 0.7 (0.6, 1.2) 0.6 (0.4, 1.1)
Physicians × 1,000 0.1 (0.0, 0.3) 0.1 (0.0, 0.2) 0.1 (0.1, 0.1) 0.1 (0.1, 0.1) 0.1 (0.0, 0.1)
Hand washing (%) 15.2 (11.8, 16.3) 7.7 (6.2, 13.4) 13.2 (8.8, 19.0) 9.9 (7.8, 23.4) 11.3 (6.6, 16.4)
Safe water drinking (%) 67.2 (48.2, 77.2) 52.7 (43.7, 62.0) 57.3 (52.2, 64.1) 60.8 (52.3, 67.1) 57.2 (47.7, 67.4)
Open defecation (%) 30.1 (17.6, 52.4) 22.0 (4.4, 42.9) 23.3 (8.5, 27.2) 37.7 (22.7, 58.2) 24.1 (13.2, 47.7)
Exclusive breastfeeding (%) 34.5 (27.8, 45.7) 38.2 (15.4, 56.7) 20.7 (11.0, 31.2) 32.7 (22.0, 47.1) 32.6 (17.3, 47.0)
Animal fat (kg/capita/year) 0.8 (0.5, 1.3) 0.4 (0.3, 0.7) 0.4 (0.3, 0.8) 0.4 (0.1, 2.0) 0.5 (0.3, 0.8)
Cereals (kg/capita/year) 152 (134, 170) 117 (98.2, 140) 127 (64.4, 152) 109 (105, 114) 127 (104, 151)
Eggs (kg/capita/year) 1.7 (1.0, 2.2) 1.5 (0.8, 2.7) 0.6 (0.5, 0.8) 1.5 (0.6, 2.1) 1.1 (0.6, 1.8)
Fish (kg/capita/year) 9.2 (3.4, 23.5) 7.0 (4.7, 13.2) 6.9 (3.0, 13.3) 5.0 (3.1, 19.8) 7.5 (3.5, 14.9)
Fruit (kg/capita/year) 23.4 (11.4, 60.2) 48.0 (29.5, 152) 19.6 (10.2, 52.3) 37.5 (24.2, 44.8) 44.8 (17.1, 71.7)
Meat (kg/capita/year) 15.6 (9.7, 24.3) 14.1 (11.3, 26.9) 19.2 (13.0, 25.7) 9.1 (7.7, 23.2) 14.6 (10.8, 22.4)
Milk (kg/capita/year) 57.1 (25.4, 91.2) 21.0 (8.5, 35.6) 17.4 (12.4, 22.6) 5.7 (4.6, 89.1) 20.5 (9.4, 44.9)
Starchy roots (kg/capita/year) 20.9 (11.9, 59.0) 156 (93.4, 187) 200 (49.8, 255) 98.9 (58.5, 236) 114 (45.4, 204)
Vegetable oils (kg/capita/year) 11.6 (5.1, 15.0) 6.3 (3.8, 8.4) 9.0 (4.6, 11.4) 9.5 (7.6, 12.4) 8.4 (5.0, 12.3)
Vegetables (kg/capita/year) 45.3 (23.6, 57.5) 29.0 (22.7, 42.8) 23.7 (21.0, 37.4) 35.1 (17.6, 47.1) 31.7 (21.1, 47.7)
Basic determinants of child mortality
GNI × capita (US $ capita) 620 (340, 1970) 500 (300, 810) 650 (410, 1350) 630 (390, 1320) 590 (360, 1150)
Unemployment rate (%) 8.1 (5.7, 21.9) 5.8 (3.5, 6.6) 4.6 (4.0, 7.5) 5.2 (1.9, 22.2) 6.0 (3.7, 10.0)
Literacy rate (%) 52.5 (41.9, 88.7) 51.2 (28.7, 65.9) 46.3 (33.6, 67.9) 61.4 (34.7, 81.7) 54.8 (34.7, 71.3)
GDP education (%) 5.1 (3.7, 6.1) 3.2 (2.4, 4.6) 2.7 (2.3, 3.5) 4.4 (3.5, 6.1) 3.7 (2.7, 5.1)
Open in a new tab

Note. GDP: gross domestic product; GNI: gross national income.

A single latent variable for groups of variables associated to a given latent factor (e.g., the availability of hand washing and safe drinking water facilities and number of people practicing open defecation to represent environment hygiene latent factor) was performed by means of principal component analysis. Five SEMs were performed considering correlated errors and multiple association paths among variables. Graphical representation was provided using the common standards where latent factors were reported as circles, instrumental variables were reported as rectangular, and arrow with the direction from the covariate to the outcomes were used to portray the paths of associations.

First, a model considering the full latent factors framework as predictors and under‐five mortality as outcome was interpolated (Figure 1a). A partial least square analysis was here used to validate the latent factor to latent factor steps of this framework confirming that one single latent factor was sufficient to represent the variables. Afterwards, immediate, underlying, and basic determinants of under‐five mortality were interpolated separately as reported on Figure 1b–d. The module of the standardized beta coefficients resulting from the SEM were reported on the arrows describing the relation between predictors and outcomes. Finally, a comprehensive SEM, considering all variables in a single framework, was interpolated separately. Supplementary analyses defined according to main founding were performed. First, the role of food resources was further investigated excluding breastfeeding from the main model. Second, a separate model was performed considering child mortality in the first year of age because this represents about the 70% of the under‐five years mortality (UNICEF webpage mortality data, March 2018).

Figure 1.

Figure 1

Open in a new tab

Structural equation analysis of child mortality. (a) Overall latent factor framework. (b–d) Immediate, underlying, and basic determinants of child mortality

Borderline significant and statistically significant paths of associations were reported on Table 2. Data management and statistical analysis were performed using the SAS statistical software package version 9.4. Borderline significant and statistically significant coefficients were defined by a type‐I error rate of 10% and 5%, respectively (α = 0.10 and 0.05).

Table 2.

Borderline and statistically significant paths of association among variables

Outcome Predictor |βst| (95% CI) P value
Immediate determinants
Child mortality ← Wasting 0.28 [−0.02, 0.58] 0.0657
Child mortality ← Underweight 0.27 [−0.03, 0.57] 0.0821
Child mortality ← Low birthweight 0.42 [0.15, 0.69] 0.0024
Underlying determinants
Wasting ← Exclusive breastfeeding 0.28 [−0.03, 0.59] 0.0755
Underweight ← Eggs availability 0.28 [−0.01, 0.58] 0.0621
Underweight ← Meat availability 0.29 [−0.01, 0.59] 0.0583
Underweight ← Exclusive breastfeeding 0.34 [0.07, 0.61] 0.0123
Wasting ← Health expenditure 0.35 [0.10, 0.61] 0.0072
Wasting ← Hospital beds 0.28 [0.01, 0.55] 0.0445
Low birthweight ← Hospital beds 0.45 [0.19, 0.71] 0.0007
Low birthweight ← Nurses and midwives 0.39 [0.13, 0.65] 0.0035
Low birthweight ← Physicians 0.29 [0.01, 0.58] 0.0413
Wasting ← Open defecation 0.43 [0.16, 0.70] 0.0020
Underweight ← Safe water drinking 0.45 [0.21, 0.69] 0.0003
Low birthweight ← Open defecation 0.27 [−0.01, 0.56] 0.0599
Low birthweight ← Safe water drinking 0.30 [0.01, 0.60] 0.0434
Basic determinants
Animal fat availability ← Unemployment rate 0.47 [0.13, 0.80] 0.0060
Cereals availability ← Unemployment rate 0.51 [0.13, 0.90] 0.0094
Cereals availability ← GDP % education 0.44 [0.08, 0.80] 0.0166
Eggs availability ← GNI × capita 0.12 [−0.01, 0.26] 0.0761
Eggs availability ← Unemployment rate 0.21 [−0.03, 0.44] 0.0892
Eggs availability ← Literacy rate 0.19 [0.07, 0.31] 0.0015
Fish availability ← Unemployment rate 0.11 [−0.02, 0.23] 0.0878
Fruit availability ← Literacy rate 0.35 [0.01, 0.68] 0.0455
Fruit availability ← Literacy rate 0.21 [0.05, 0.38] 0.0091
Meat availability ← Literacy rate 0.27 [0.12, 0.41] 0.0003
Milk availability ← Literacy rate 0.38 [0.18, 0.59] 0.0003
Starchy roots availability ← GDP % education 0.29 [0.07, 0.51] 0.0085
Starchy roots availability ← Unemployment rate 0.68 [0.48, 0.88] 0.0021
Exclusive breastfeeding ← GNI × capita 0.12 [0.00, 0.25] 0.0544
Exclusive breastfeeding ← Unemployment rate 0.27 [0.05, 0.49] 0.0181
Exclusive breastfeeding ← Literacy rate 0.17 [0.06, 0.27] 0.0023
Hospital beds ← GNI × capita 0.28 [0.13, 0.43] 0.0002
Nurses and midwives ← GNI × capita 0.18 [−0.02, 0.39] 0.0786
Nurses and midwives ← Unemployment rate 0.44 [0.08, 0.81] 0.0164
Nurses and midwives ← Literacy rate 0.24 [0.07, 0.41] 0.0062
Physicians ← Unemployment rate 0.53 [0.22, 0.84] 0.0007
Open defecation ← GDP % education 0.18 [−0.03, 0.39] 0.0855
Safe water drinking ← GNI × capita 0.12 [0.00, 0.25] 0.0539
Open in a new tab

Note. GDP: gross domestic product; GNI: gross national income.

The Proc PRINCOMP and the Proc partial least square analysis of the SAS statistical software package were used to perform latent factors and to validate the latent factor to latent factor steps of the comprehensive latent factor framework model (Figure 1a). Finally, the Proc CALIS of the SAS statistical software package was used to interpolate the SEMs.

3. RESULTS

Under‐five mortality in sub‐Saharan Africa during the period 2000–2016 ranged between 76.4 × 1,000 and 127 × 1,000 as observed in Eastern and Central Africa, respectively. Underweight prevalence ranged between 11.6% and 21.1% for Southern and Western sub‐Saharan Africa. Stunting prevalence ranged between 32.8% and 44.3% for Southern and Eastern sub‐Saharan Africa. Finally, low birthweight ranged between 11.4% and 16.0% for Southern and Western sub‐Saharan Africa, respectively. Underlying and basic determinants of child mortality are summarized in Table 1.

When looking at our latent factor framework, we observed that nutritional status and child mortality are significantly correlated, having the module of the standardized beta coefficient (|βst|) of 0.43 (95% CI [0.17, 0.70]). In this framework, care resources, hygiene, and sanitation are, in turn, strongly related to health status, having a |βst| of 0.57 [0.35, 0.79] and 0.60 [0.39, 0.81], respectively. Finally, poverty is strongly associated to care resources availability (|βst| = 0.76 [0.60, 0.91]) but not with other underlying factors of child mortality. On the other hand, among immediate determinants of child mortality, low birthweight rate is significantly associated to under‐five mortality (|βst| = 0.42 [0.15, 0.69]), whereas the association between stunting and underweight is weaker and borderline significant (|βst| = 0.28 [−0.02, 0.58] and 0.27 [−0.03, 0.57], respectively). When looking at underlying factors related to child mortality, we observed that exclusive breastfeeding (|βst| = 0.34 [0.05, 0.63]), but not any other food availability factors, has a significant association with child nutritional status. Among other underlying factors of child nutritional status, we observed that all of the variables representing health care resources availability were associated to child nutritional status (|βst| = 0.53 [0.30, 0.77], 0.64 [0.45, 0.83], 0.50 [0.25, 0.74], and 0.26 [−0.05, 0.56] for health expenditure, hospital beds, number of nurses and midwives, and number of physicians, respectively). Among environmental hygiene factors, open defecation practice and availability of safe drinking water emerged as significantly associated with child nutritional status (|βst| = 0.61 [0.40, 0.82] and 0.38 [0.10, 0.66] for open defecation practice and availability of safe drinking water, respectively). Finally, among basic determinants of child mortality, we observed a significant association between gross national income, unemployment, and literacy rates with health resources availability (|βst| = 0.84 [0.71, 0.97], 0.51 [0.24, 0.79], and 0.66 [0.47, 0.86], respectively). On the other hand, we observed a moderate borderline significant association between environmental sanitation and hygiene with education expenditure (|βst| = 0.29 [−0.05, 0.62]).

When looking at the SEM considering all of the variables, we observed that most of the above‐reported paths of associations are confirmed, whereas others emerged. Among those, we observed a borderline to significant association between wasting and underweight with exclusive breastfeeding (|βst| = 0.28 [−0.03, 0.59] and 0.34 [0.07, 0.61], respectively). Underweight, but not wasting, was also borderline associated to animal protein intakes in the forms of eggs and meat availabilities (|βst| = 0.28 [−0.01, 0.58] and 0.29 [−0.01, 0.59], respectively). On the other side, it is interesting to notice that low birthweight and wasting, but not underweight, were significantly associated with all of those variables representing health care resources with |βst| ranging between 0.28 and 0.39. Moreover, wasting and underweight prevalence and low birthweight rate were associated to sanitation and hygiene factors as open defecation practice and availability of safe drinking water with |βst| ranging between 0.27 and 0.45. Basic determinants of child mortality as expenditure on education, unemployment, and literacy rates were associated to food variables such as eggs, meat, fish, and starchy roots availabilities with |βst| ranging between 0.19 and 0.68. Exclusive breastfeeding was associated with unemployment and literacy rates (|βst| = 0.27 [0.05, 0.49] and 0.17 [0.06, 0.27], respectively). Finally, those basic determinants of child mortality as gross national income, literacy, and unemployment rates were significantly associated to health resources (|βst| ranging between 0.24 and 0.53) and just borderline significantly associated with hygiene and sanitation variables (|βst| ranging between 0.12 and 0.18).

3.1. Supplementary analyses

When considering child mortality in the first year of age, the present results are substantially confirmed. In particular, it was observed that the role of breastfeeding in determining the child nutritional status was even reinforced so that the |βst| coefficient raised up from 0.34 to 0.55. Not surprisingly, we observed that low birthweight rate was much stronger associated to child mortality in the first year of age. As a consequence, the |βst| coefficient raised up from 0.42 to 0.63.

4. DISCUSSION

In the present work, a conceptual framework of the under‐five mortality and nutritional status in sub‐Saharan Africa was discussed from a quantitative viewpoint. Thus, knowledge about child mortality has been reinforced, and relations between factors determining child mortality and health were clarified and assessed quantitatively. First, it appeared that low birthweight, more than underweight or stunting, influences child mortality. Similarly, low birthweight and child mortality has been consistently associated in studies conducted in Africa and other low‐ to middle‐income countries (Cooper & Sandler, 1997; Katz et al., 2013; Marchant et al., 2012). Moreover, it has been widely reported that the nutritional status of the mother and the birthweight of the child are strongly related (Bergner & Susser, 1970; Kramer, 1987). This suggests that improving the nutritional status of the mother would indirectly result in a reduction in child mortality probably through a reduction in low birthweight rate. It was observed that breastfeeding, more than food availability, can be related to child nutritional status, confirming the importance of interventions aimed to encourage this practice (Tylleskär et al., 2011). This result can be explained by the fact that up to 70% of child mortality occurs in the first year of age when exclusive breastfeeding is particularly important to determine child development and health. Exclusive breastfeeding practice in children up to 6 months of age was estimated to be 35% in Africa in 2010, whereas in Asia and other low‐ to middle‐income countries, it was estimated as 41% and 39%, respectively (Cai, Wardlaw, & Brown, 2012). The existing difference in exclusive breastfeeding practices between Africa and other low‐ to middle‐income countries may have affected the nutritional status of the children resulting in the observed higher child mortality rates in Africa compared with other low‐ to middle‐income areas (Liu et al., 2015; You et al., 2015). Nevertheless, exclusive breastfeeding during early stages of life, more than nutrition after weaning, is important in influencing the health status of children and should be implemented (Kimani‐Murage et al., 2011; Lartey, 2008; Sankar et al., 2015; Vaahtera et al., 2001) .

It is worth noting that breastfeeding improves the health status of children through a mechanism of improved immune resistance, as reported for HIV and other infectious diseases (Sankar et al., 2015; Victoria, 2000; Wood, Wiseman, Morales, Gedamke, & Castro, 2000). In the present work, the absence of a significant relation between food availability and nutritional status remain a challenge. This unexpected negative result could be attributed to at least two different reasons. First, the limited sample size, which refers to 37 countries only, could not be sufficient to detect significant small associations leading to a type‐II error. Second, it could be due to the fact that, among the food availability factors, exclusive breastfeeding was the strongest and captured the most of the nutritional status variability. This second hypothesis is confirmed by a sensitivity analysis performed without considering exclusive breastfeeding. According to this analysis, significant associations between foods rich in animal proteins (eggs and meat but not fish) and low birthweight, underweight, and stunting emerged. Furthermore, the correlation among exclusive breastfeeding and other food sources was not statistically significant, confirming that breastfeeding is not only the main nutritional determinant of child mortality, but it is also quite independent from food intakes. Again, this is likely due to the fact that the vast majority of child mortality take place in the first year of age.

The present work confirm that there is a relation between environmental hygiene, in the form of safe drinking water and practice of open defecation, and the nutritional status of children, defined as underweight, wasting, and low birthweight. This association is well acknowledged and documented and is likely driven by diarrhoea or related infectious diseases owing to poor hygienic conditions in sub‐Saharan Africa (Oloruntoba, Folarin, & Ayede, 2014; Westaway & Viljoen, 2000).

Nonetheless, a significant relation between poverty and health care resources that, in turn, are related to the nutritional status of the child was observed. This may confirm that poverty is a basic factor of child nutritional status and under‐five mortality as well (Cairncross, Bartram, Cumming, & Brocklehurst, 2010; Garriga & Foguet, 2013). Finally, among factors related to poverty, education was confirmed to be related to health resources, hygiene and sanitation, and exclusive breastfeeding. This evidence reinforces the idea that improvement in education may enhance hygiene conditions and, in effect, child health status (Barrett & Browne, 1996; Browne & Barrett, 1991).

The present work confirms and reinforces the statements defined by the Sustainable Development Goals (SDG) to improve the status of sub‐Saharan Africa (World Health Organization, 2004). Among these goals, eradication of poverty, ensuring inclusive and quality education for all, and reduction of child mortality are connected. In particular, the present work show a hierarchical relation among these SDGs, suggesting that acting on reducing poverty at the base of the chain may positively influence child mortality. According to this study, these mechanisms act through the improvement of nutritional status of children showing how nutritional status is an important determinant of child mortality. SDGs has been criticized in the past (Attaran, 2005; Easterly, 2009). Among main criticisms, it was reported that the SDGs are not directly or easily measurable. Therefore, this study aimed to evaluate how the factors under investigation influenced under‐five mortality and how they are correlated.

The main strength of the current work is that it was performed by means of an integrated system aimed to evaluate multiple relations from a quantitative viewpoint. Moreover, the analysis was based on reliable data defined over a wide observational period. Finally, a comprehensive panel of data from 37 countries, well representing all sub‐Saharan Africa, was considered, making the present work fairly generalizable to all sub‐Saharan Africa as a whole. Moreover, in the present work, we report about a hierarchy among factors concurring in under‐five child mortality. This hierarchy could be useful to address policymakers and stakeholder in taking decisions aimed to reduce the burden of child mortality in sub‐Saharan Africa. For example, among immediate determinant of child mortality, we confirmed that acting on low birthweight is a priority (Lubchenco, Searls, & Brazie, 1972; McCormick, 1985). When considering underlying determinants of child mortality, we reported that improved hygiene by reducing open defecation could be a possible solution. Notably, open defecation prevalence is still a main concern in sub‐Saharan Africa, where in certain areas, it could exceed 50% (Galan, Kim, & Graham, 2013). Nevertheless, it was reported that acting on open defecation is a cost‐effective strategy to reduce diarrhoea and infections that, in turn, are related to child mortality (Clasen et al., 2014; Pickering, Djebbari, Lopez, Coulibaly, & Alzua, 2015). In the present work, we reported that improving health care facilities could result as more effective than acting on other components in agreement with the need of guaranteeing a continuum of care between mother and the newborn. (Group, 2008; Kerber et al., 2007; Tinker, ten Hoope‐Bender, Azfar, Bustreo, & Bell, 2005). In this context, any effort made to enhance exclusive breastfeeding is positive as well (Edmond et al., 2006). Finally, the present work highlights the important role of underlying determinant of child mortality as literacy rates and economic income (Browne & Barrett, 1991; Wagstaff, 2000).

The present work is not free from any limitations. Here, we supposed a defined structure for our SEMs, where the relations between variables are defined in the form of hierarchy, which in turn is defined according to a well‐accepted conceptual framework. This hierarchy defined the causation, introducing a strong a priori assumption. This assumption is likely but not necessarily true. We here focused on just a subset of the determinants of child mortality, whereas others such as HIV, tuberculosis, and other infectious diseases remain unexplored. On the other hand, the scope of the present work was mainly focused on nutrition, an easier modifiable determinant of child mortality. Furthermore, the strength of the causation could be improved considering the time of exposure to the predictor and the sequentiality of the outcome. This approach could not be performed due to the limitations of the SEM analyses, where models accounting for variations over the time are just poorly implemented.

Many of the proposed associations are well known and extensively reported in the literature. The present work confirms and reinforces such evidence using a much more comprehensive approach aimed to consider child mortality and its determinants as a whole. The authors agree that the present work may represent just an early attempt to a much more integrated investigation of the determinants of child mortality in sub‐Saharan Africa. We prospect that further similar studies will be conducted because the complex relations among the factors concurring in child mortality in sub‐Saharan Africa is not yet fully understood.

CONFLICTS OF INTEREST

The authors declare that they have no conflicts of interest.

CONTRIBUTIONS

CR conceptualized the study, collected the data, performed the statistical analysis, and drafted the manuscript; JC and HA participated to data collection and contributed to the first version of the manuscript; CMS furnished logistical and technical support; ML and RD supervised the work and furnished technical support and nutritional advices. All authors contributed revising the final version of the manuscript and approved it.

Supporting information

Table S1. List of variables and their definitions.

Click here for additional data file. (14.9KB, docx)

ACKNOWLEDGMENT

The authors sincerely thank Mrs. Ronel Benson and Mrs. Henriette Claaseen for administrative support furnished.

Ricci C, Carboo J, Asare H, Smuts CM, Dolman R, Lombard M. Nutritional status as a central determinant of child mortality in sub‐Saharan Africa: A quantitative conceptual framework. Matern Child Nutr. 2019;15:e12722 10.1111/mcn.12722

REFERENCES

  1. Attaran, A. (2005). An immeasurable crisis? A criticism of the Millennium Development Goals and why they cannot be measured. PLoS Medicine, 2(10), e318 10.1371/journal.pmed.0020318 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Barrett, H. , & Browne, A. (1996). Health, hygiene and maternal education: Evidence from The Gambia. Social Science & Medicine, 43(11), 1579–1590. 10.1016/S0277-9536(96)00054-8 [DOI] [PubMed] [Google Scholar]
  3. Bergner, L. , & Susser, M. W. (1970). Low birth weight and prenatal nutrition: an interpretative review. Pediatrics, 46(6), 946–966. [PubMed] [Google Scholar]
  4. Black, R. E. , Morris, S. S. , & Bryce, J. (2003). Where and why are 10 million children dying every year? The Lancet, 361(9376), 2226–2234. 10.1016/S0140-6736(03)13779-8 [DOI] [PubMed] [Google Scholar]
  5. Blom, G. (1958). Statistical estimates and transformed beta variables. New York: Wiley. [Google Scholar]
  6. Browne, A. W. , & Barrett, H. R. (1991). Female education in sub‐Saharan Africa: The key to development? Comparative Education, 27(3), 275–285. 10.1080/0305006910270303 [DOI] [Google Scholar]
  7. Cai, X. , Wardlaw, T. , & Brown, D. W. (2012). Global trends in exclusive breastfeeding. International Breastfeeding Journal, 7(1), 12 10.1186/1746-4358-7-12 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Cairncross, S. , Bartram, J. , Cumming, O. , & Brocklehurst, C. (2010). Hygiene, sanitation, and water: What needs to be done? PLoS Medicine, 7(11), e1000365 10.1371/journal.pmed.1000365 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Clasen, T. , Boisson, S. , Routray, P. , Torondel, B. , Bell, M. , Cumming, O. , … Odagiri, M. (2014). Effectiveness of a rural sanitation programme on diarrhoea, soil‐transmitted helminth infection, and child malnutrition in Odisha, India: A cluster‐randomised trial. The Lancet Global Health, 2(11), e645–e653. 10.1016/S2214-109X(14)70307-9 [DOI] [PubMed] [Google Scholar]
  10. Cooper, P. A. , & Sandler, D. L. (1997). Outcome of very low birth weight infants at 12 to 18 months of age in Soweto, South Africa. Pediatrics, 99(4), 537–544. [DOI] [PubMed] [Google Scholar]
  11. Easterly, W. (2009). How the millennium development goals are unfair to Africa. World Development, 37(1), 26–35. 10.1016/j.worlddev.2008.02.009 [DOI] [Google Scholar]
  12. Edmond, K. M. , Zandoh, C. , Quigley, M. A. , Amenga‐Etego, S. , Owusu‐Agyei, S. , & Kirkwood, B. R. (2006). Delayed breastfeeding initiation increases risk of neonatal mortality. Pediatrics, 117(3), e380–e386. 10.1542/peds.2005-1496 [DOI] [PubMed] [Google Scholar]
  13. Engle, P. L. , Menon, P. , & Haddad, L. (1999). Care and nutrition: Concepts and measurement. World Development, 27(8), 1309–1337. 10.1016/S0305-750X(99)00059-5 [DOI] [Google Scholar]
  14. Foster, G. , & Williamson, J. (2000). A review of current literature on the impact of HIV/AIDS on children in sub‐Saharan Africa ( ed., Vol. 14) (pp. S275–S284)Aids‐London‐Current Science Then Rapid Science Publishers Then Lippincott Raven. [PubMed] [Google Scholar]
  15. Galan, D. I. , Kim, S.‐S. , & Graham, J. P. (2013). Exploring changes in open defecation prevalence in sub‐Saharan Africa based on national level indices. BMC Public Health, 13(1), 527 10.1186/1471-2458-13-527 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Garriga, R. G. , & Foguet, A. P. (2013). Water, sanitation, hygiene and rural poverty: Issues of sector monitoring and the role of aggregated indicators. Water Policy, 15(6), 1018–1045. 10.2166/wp.2013.037 [DOI] [Google Scholar]
  17. Group, S. A. E. D. C. W. (2008). Every death counts: Use of mortality audit data for decision making to save the lives of mothers, babies, and children in South Africa. The Lancet, 371(9620), 1294–1304. [DOI] [PubMed] [Google Scholar]
  18. Jamison, D. T. (2006). Disease and mortality in sub‐Saharan Africa. World Bank Publications. [PubMed] [Google Scholar]
  19. Katz, J. , Lee, A. C. , Kozuki, N. , Lawn, J. E. , Cousens, S. , Blencowe, H. , … Willey, B. A. (2013). Mortality risk in preterm and small‐for‐gestational‐age infants in low‐income and middle‐income countries: A pooled country analysis. The Lancet, 382(9890), 417–425. 10.1016/S0140-6736(13)60993-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Kerber, K. J. , de Graft‐Johnson, J. E. , Bhutta, Z. A. , Okong, P. , Starrs, A. , & Lawn, J. E. (2007). Continuum of care for maternal, newborn, and child health: From slogan to service delivery. The Lancet, 370(9595), 1358–1369. 10.1016/S0140-6736(07)61578-5 [DOI] [PubMed] [Google Scholar]
  21. Kimani‐Murage, E. W. , Madise, N. J. , Fotso, J.‐C. , Kyobutungi, C. , Mutua, M. K. , Gitau, T. M. , & Yatich, N. (2011). Patterns and determinants of breastfeeding and complementary feeding practices in urban informal settlements, Nairobi Kenya. BMC Public Health, 11(1), 396 10.1186/1471-2458-11-396 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Kramer, M. S. (1987). Determinants of low birth weight: Methodological assessment and meta‐analysis. Bulletin of the World Health Organization, 65(5), 663. [PMC free article] [PubMed] [Google Scholar]
  23. Lartey, A. (2008). Maternal and child nutrition in sub‐Saharan Africa: Challenges and interventions. Proceedings of the Nutrition Society, 67(1), 105–108. 10.1017/S0029665108006083 [DOI] [PubMed] [Google Scholar]
  24. Lim, S. S. , Vos, T. , Flaxman, A. D. , Danaei, G. , Shibuya, K. , Adair‐Rohani, H. , … Andrews, K. G. (2012). A comparative risk assessment of burden of disease and injury attributable to 67 risk factors and risk factor clusters in 21 regions, 1990–2010: A systematic analysis for the Global Burden of Disease Study 2010. The Lancet, 380(9859), 2224–2260. 10.1016/S0140-6736(12)61766-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Liu, L. , Oza, S. , Hogan, D. , Perin, J. , Rudan, I. , Lawn, J. E. , … Black, R. E. (2015). Global, regional, and national causes of child mortality in 2000–13, with projections to inform post‐2015 priorities: An updated systematic analysis. The Lancet, 385(9966), 430–440. 10.1016/S0140-6736(14)61698-6 [DOI] [PubMed] [Google Scholar]
  26. Lubchenco, L. O. , Searls, D. , & Brazie, J. (1972). Neonatal mortality rate: Relationship to birth weight and gestational age. The Journal of Pediatrics, 81(4), 814–822. 10.1016/S0022-3476(72)80114-8 [DOI] [PubMed] [Google Scholar]
  27. Marchant, T. , Willey, B. , Katz, J. , Clarke, S. , Kariuki, S. , Ter Kuile, F. , … Watson‐Jones, D. (2012). Neonatal mortality risk associated with preterm birth in East Africa, adjusted by weight for gestational age: Individual participant level meta‐analysis. PLoS Medicine, 9(8), e1001292 10.1371/journal.pmed.1001292 [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. McCormick, M. C. (1985). The contribution of low birth weight to infant mortality and childhood morbidity. New England Journal of Medicine, 312(2), 82–90. 10.1056/NEJM198501103120204 [DOI] [PubMed] [Google Scholar]
  29. Mosley, W. H. (1983). Will primary health care reduce infant and child mortality? A critique of some current strategies with special reference to Africa and Asia [draft].
  30. Oloruntoba, E. O. , Folarin, T. B. , & Ayede, A. I. (2014). Hygiene and sanitation risk factors of diarrhoeal disease among under‐five children in Ibadan, Nigeria. African health sciences, 14(4), 1001–1011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. WHO (2004). Millennium development goals. WHO regional Office for South‐East Asia. [Google Scholar]
  32. Pelletier, D. L. (1994). The potentiating effects of malnutrition on child mortality: Epidemiologic evidence and policy implications. Nutrition Reviews, 52(12), 409–415. [DOI] [PubMed] [Google Scholar]
  33. Pickering, A. J. , Djebbari, H. , Lopez, C. , Coulibaly, M. , & Alzua, M. L. (2015). Effect of a community‐led sanitation intervention on child diarrhoea and child growth in rural Mali: A cluster‐randomised controlled trial. The Lancet Global Health, 3(11), e701–e711. 10.1016/S2214-109X(15)00144-8 [DOI] [PubMed] [Google Scholar]
  34. Sankar, M. J. , Sinha, B. , Chowdhury, R. , Bhandari, N. , Taneja, S. , Martines, J. , & Bahl, R. (2015). Optimal breastfeeding practices and infant and child mortality: A systematic review and meta‐analysis. Acta Paediatrica, 104(S467), 3–13. 10.1111/apa.13147 [DOI] [PubMed] [Google Scholar]
  35. Tinker, A. , ten Hoope‐Bender, P. , Azfar, S. , Bustreo, F. , & Bell, R. (2005). A continuum of care to save newborn lives. The Lancet, 365(9462), 822–825. 10.1016/S0140-6736(05)71016-3 [DOI] [PubMed] [Google Scholar]
  36. Tylleskär, T. , Jackson, D. , Meda, N. , Engebretsen, I. M. S. , Chopra, M. , Diallo, A. H. , … Goga, A. (2011). Exclusive breastfeeding promotion by peer counsellors in sub‐Saharan Africa (PROMISE‐EBF): A cluster‐randomised trial. The Lancet, 378(9789), 420–427. 10.1016/S0140-6736(11)60738-1 [DOI] [PubMed] [Google Scholar]
  37. UNICEF webpage mortality data . (2018). from https://data.unicef.org/topic/child-survival/under-five-mortality/
  38. Vaahtera, M. , Kulmala, T. , Hietanen, A. , Ndekha, M. , Cullinan, T. , Salin, M. L. , & Ashorn, P. (2001). Breastfeeding and complementary feeding practices in rural Malawi. Acta Paediatrica, 90(3), 328–332. 10.1111/j.1651-2227.2001.tb00313.x [DOI] [PubMed] [Google Scholar]
  39. Victoria, C. (2000). Effect of breastfeeding on infant and child mortality due to infectious diseases in less developed countries: A pooled analysis. Lancet (British Edition), 355(9202), 451–455. [PubMed] [Google Scholar]
  40. Wagstaff, A. (2000). Socioeconomic inequalities in child mortality: Comparisons across nine developing countries. Bulletin of the World Health Organization, 78, 19–29. [PMC free article] [PubMed] [Google Scholar]
  41. Westaway, M. S. , & Viljoen, E. (2000). Health and hygiene knowledge, attitudes and behaviour. Health & place, 6(1), 25–32. 10.1016/S1353-8292(99)00027-1 [DOI] [PubMed] [Google Scholar]
  42. Wood, B. , Wiseman, K. , Morales, L. , Gedamke, K. M. , & Castro, C. (2000). Effect of breastfeeding on infant and child mortality due to infectious diseases in less developed countries: A pooled analysis. WHO collaborative study team on the role of breastfeeding on the prevention of infant mortality. The Lancet, 355(9202), 451–455. [PubMed] [Google Scholar]
  43. You, D. , Hug, L. , Ejdemyr, S. , & Beise, J. (2015). Levels and trends in child mortality: Report 2015. Estimates developed by the UN Inter‐agency Group for Child Mortality Estimation.

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Table S1. List of variables and their definitions.

Click here for additional data file. (14.9KB, docx)

Articles from Maternal & Child Nutrition are provided here courtesy of Wiley

RESOURCES