Socio-Economic Inequalities in the Double Burden of Malnutrition among under-Five Children: Evidence from 10 Selected Sub-Saharan African Countries

Background: Africa is unlikely to end hunger and all forms of malnutrition by 2030 due to public health problems such as the double burden of malnutrition (DBM). Thus, the aim of this study is to determine the prevalence of DBM and degree of socio-economic inequality in double burden of malnutrition among children under 5 years in sub-Saharan Africa. Methods: This study used multi-country data collected by the Demographic and Health Surveys (DHS) Program. Data for this analysis were drawn from the DHS women’s questionnaire focusing on children under 5 years. The outcome variable for this study was the double burden of malnutrition (DBM). This variable was computed from four indicators: stunting, wasting, underweight and overweight. Inequalities in DBM among children under 5 years were measured using concentration indices (CI). Results: The total number of children included in this analysis was 55,285. DBM was highest in Burundi (26.74%) and lowest in Senegal (8.80%). The computed adjusted Erreygers Concentration Indices showed pro-poor socio-economic child health inequalities relative to the double burden of malnutrition. The DBM pro-poor inequalities were most intense in Zimbabwe (−0.0294) and least intense in Burundi (−0.2206). Conclusions: This study has shown that across SSA, among under-five children, the poor suffer more from the DBM relative to the wealthy. If we are not to leave any child behind, we must address these socio-economic inequalities in sub-Saharan Africa.


Introduction
According to the World Food Programme (WFP), malnutrition is "a state in which the physical function of an individual is impaired to the point where he or she can no longer maintain adequate bodily performance process such as growth, pregnancy, lactation, physical work and resisting and recovering from disease" [1]. Malnutrition can also be termed as an abuse of food or bad nutrition, such as over-nutrition and under-nutrition. Hunger and protein-energy malnutrition (PEM) have led to high mortality rates in children and mothers, contributing to poor growth and a rise in opportunistic infections and The underlying causes of DBM vary by sub-region in sub-Saharan Africa. For example, one study found that cultural perceptions such as a heavier body size in females may signify wealth, good, stable marital home, and exceptional achievement [33]. However, these perceptions differ across sub-regions as some regions expect women to work hard and have increased physical activity. Another study attributed obesity and increased weight to the rise in consumption of cheap, processed food at the expense of fresh, non-processed foods of subsistence farming [34]. The rise in the commercialization of food production is correlated to the decrease in subsistence farming. This has led to household diets with low nutritional value, high sugar, high fat, and energy-dense food that leads to obesity [35]. This article uses a population-based study on ten African countries (Burundi, Ethiopia, Guinea, Malawi, Mali, Senegal, Sierra Leone, South Africa, Zambia, and Zimbabwe) to understand the socio-economic inequalities in the double burden malnutrition among under-5 children in the continent. Thus, the aim of this study is to determine the prevalence of DBM and degree of socio-economic inequality in double burden of malnutrition among children under 5 years in sub-Saharan Africa.

Data Source
This study used multi-country data collected by the Demographic and Health Surveys

Stunting
Children whose height-for-age Z-score was below minus three standard deviations (−3 SD) were considered stunted, and those above minus three standard deviations (−3 SD) were considered not stunted. Stunting was coded as binary variable assigned values of zero and one. Those that were not stunted above −3 SD were coded as "0", and those that were stunted below −3 SD were coded as "1".

Wasting
Children whose Z-score was below minus three standard deviations (−3 SD) from the median of the reference population were considered thin (wasted), and those above a Z-score above −3 SD were considered not wasted. Wasting was coded as binary variable assigned values of zero and one. Those that were not wasted above −3 SD were coded as "0" and those that were wasted below −3 SD were coded as "1".

Underweight
Children whose weight-for-age Z-score was below minus three standard deviations (−3 SD) from the median of the reference population were classified as underweight, and those that were above −3 SD considered not underweight. Underweight was coded as binary variable assigned values of zero and one. Those that were not underweight above −3 SD were coded as "0" and those that were underweight below −3 SD were coded as "1".

Overweight
Children whose weight-for-height Z-score was more than 2 standard deviations (+2 SD) above the median of the reference population were considered overweight. Overweight was coded as binary variable assigned values of zero and one. Those that were not overweight below +2 SD were coded as "0" and those that were overweight above +2 SD were coded as "1".

Double Burden of Malnutrition (DBM)
The outcome variable for this study is the double burden of malnutrition (DBM). This variable was computed from 4 indicators: stunting, wasting, underweight and overweight. The first step was calculating the row total of stunted, wasted, and underweight children. Then children who had a row total greater than 1 and were overweight were defined as having experienced a double burden of malnutrition. The outcome variable, double burden of malnutrition (DBM), was then recorded as a binary variable where a value of "1" was assigned if DBM was present, and a value of "0" was given if DBM was absent.

Socio-Economic Status (SES)
Socio-economic status was depicted by the household wealth index, which measures a household's cumulative standard of living, such as ownership of select assets, type of housing, sanitation services, and type of water access, among others, using Principal Component Analysis (PCA) [32]. The household wealth index is considered a more reliable measure of wealth compared to income and consumption because it reflects a household's long-term standard of living, and this makes it possible to identify problems particular to the poor members of society, such as unequal access to health care and unequal access to recommended nutrition [33]. For this study, wealth was grouped into 5 quintiles-poorest (Q1), poorer (Q2), middle (Q3), richer (Q4) and richest (Q5).

Statistical Analysis Data Analysis
The study analyzed the data using STATA 17.1 statistical software. Univariate and bivariate analyses were performed to describe the sample and patterns of DBM. Inequalities in DBM among children under 5 years were measured using concentration indices (CI). The concentration index approach is a standard measure of assessing health inequalities. The indices and curves investigate whether health inequalities exist in one group. However, they do not estimate the magnitude of health inequalities [36]. This paper used the Erreygers normalized concentration indices [37] to measure the socio-economic inequalities among children in DBM: wasting, stunting, underweight and overweight. Among many of the indices that could have been used, we opted to adopt the Normalized Erreygers Indices as they have been corrected for bound issues; hence, they give more robust standard errors. The concentration index ranges from −1 to +1 and estimates the extent to which a health outcome (DBM) is concentrated among the rich or the poor. A negative concentration index value denotes a health outcome (DBM) concentrated among the poor. In contrast, a positive value implies that a health outcome (DBM) is concentrated among the rich [36]. A concentration index of zero implies that there is no socio-economic-related inequality, and a large absolute value of the concentration index depicts a greater concentration of inequality [36,38].
The concentration index can be computed by making use of the 'covariance' as shown below: where: y i is the health variable; y is the mean of y i ; R i is the fractional rank of the ith individual; COV denotes the covariance.

Socio-Economic Inequalities
Across all countries, stunting disproportionately affected the poorest, as stunting was more prevalent in the poorest quintile (Q1) ( Table 3). The concentration indices for stunting were all negative and statistically significant at a 95% confidence interval in all countries, ranging from −0.21 in Burundi to −0.03 in Zimbabwe (Table 3). Burundi, Ethiopia, Guinea, Mali, Senegal, Sierra Leone, and Zimbabwe all reported pro-poor wasting child health inequalities ( Table 4). The latter concentration indices were statistically significant at a 95% confidence interval. While Malawi, South Africa and Zambia reported pro-rich inequalities, the concentration indices were not statistically significant at a 95% confidence interval (Table 4).
Underweight children were also more prevalent in the poorest quintile (Q1) for most of the countries except for Zimbabwe (31.43%; Q2) and South Africa (35.71; Q3): Burundi (38.84%), Ethiopia (51.52%), Guinea (30.69%), Malawi (34.11%), Mali (28.94%), Senegal (52.50%), Sierra Leone (33.33%), and Zambia (38.57%) ( Table 5). All the underweight concentration indices were statistically significant at a 95% confidence interval across countries except for South Africa which had pro-rich inequalities; otherwise, all other countries reported negative indices, indicating pro-poor underweight inequalities (Table 5).  Conversely, overweight disproportionately affected children from the rich households in many of the countries (Burundi, Ethiopia, Malawi, Mali, Senegal, Sierra Leone, Zambia and Zimbabwe); however, only Burundi, Mali, Senegal and Zimbabwe had statistically significant concentration indices at a 95% confidence interval (Table 6). While Guinea and South Africa reported pro-poor inequalities, the concentration indices were not statistically significant at a 95% confidence interval (Table 6).  (Figure 2). Across all countries, DBM was most prevalent among children in the poorest quintile (Q1) except in Zimbabwe, where DBM was most prevalent among children from the richer quintile (Q4) ( Table 7). However, the adjusted Erreygers concentration indices were negative, showing pro-poor DBM inequalities among children across all countries (Table 7).  DBM across all countries reported pro-poor inequalities as all the concentration indices were negative and statistically significant at a 95% confidence interval ( Table 8). The computed adjusted Erreygers Concentration Indices showed pro-poor socio-economic child health inequalities relative to the double burden of malnutrition. All the concentration indices were negative across all countries and were statistically significant at a 95% confidence interval ( Table 8). The DBM pro-poor inequalities were more intense in Zimbabwe (−0.0294) and least intense in Burundi (−0.2206) ( Table 8)    DBM across all countries reported pro-poor inequalities as all the concentration indices were negative and statistically significant at a 95% confidence interval ( Table 8). The computed adjusted Erreygers Concentration Indices showed pro-poor socio-economic child health inequalities relative to the double burden of malnutrition. All the concentration indices were negative across all countries and were statistically significant at a 95% confidence interval ( Table 8). The DBM pro-poor inequalities were more intense in Zimbabwe (−0.0294) and least intense in Burundi (−0.2206) (Table 8) Table 8).

Discussion
SDG Target 2.2 aims to "End all forms of malnutrition, including achieving, by 2020, the internationally agreed targets on stunting and wasting in children under 5 years of age". In this regard, our study contributes several ways to the debate on DBM among children in African countries. First, our results provide evidence on individual nutritional status (underweight, wasting, stunting, and overweight) prevalence of children under 5 years of age at the national level and explain the existence of socio-economic inequalities. The quality of evidence for our approach is supported by representative DHS data from 10 African countries. The completeness of the dataset for analysis, as this study has drawn insights from the most recent dataset from 2015 to 2019, suggests that the geographic and social differences in DBM of under-five children in Africa and the extent of economic inequality can be fully understood.
Sub-Saharan Africa has been cited to be characterized by the double burden of malnutrition (DBM) and high levels of undernutrition as well as a growing burden of overweight/obesity and diet-related non-communicable diseases (NCDs) [39]. Recent research shows that despite a high prevalence of hunger and malnutrition, overweight and obesity epidemics are increasing in Africa [40]. This is still the case, as our study findings showed a significantly high prevalence of DBM, with Senegal reporting the least DBM prevalence of about 9%. In comparison, Burundi had the highest DBM prevalence of about 27%. A recent study raised a concern, citing the possibility that most countries will not meet the global nutrition targets by 2030 [39] and Africa is unlikely to reach the Sustainable Development Goals and end hunger and all forms of malnutrition by 2030. The current study findings seem to show the concern raised in earlier papers becoming a sad reality as this study reported a significantly high prevalence of stunting (Burundi; 24%), underweight (Burundi; 8%) and overweight (South Africa; 13%). Earlier research reported malnutrition commonly observed in developed and affluent communities, but as early as 1996, it was noticed in low-to-middle-income countries (LMICs) [41][42][43].
A recent study reported the prevalence of overweight and obesity among under-five children in South Africa to be almost double that of Malawi [44]. Our results also showed similar findings: South Africa had the highest overweight prevalence of 13% compared to Senegal, which had about 2%. However, it had the 5th highest DBM prevalence of 16%, with Burundi with the highest DBM prevalence of 27%. The reported that the high DBM prevalence of Burundi could be attributed to earlier trends of stunting and underweight among children, which have shown little to no changes since the 1980s [43].
Contrary to what was observed in previous studies that reported relatively socioeconomic well-off groups at a greater risk for the double burden of malnutrition [44][45][46], our study showed that DBM was more prevalent among the children from the poorest households (Q1). This may be because of a shift in NCDs' epidemiology, as they were earlier perceived as diseases for developed countries but are currently more prevalent in developing countries. Furthermore, a DBM is a global problem. It has been argued that it occurs when the prevalence of overweight and obesity in LMICs is increasing rapidly, while at the same time, the prevalence of malnutrition in these countries is declining slowly [47]. This was true for our study, as Burundi had the highest prevalence of stunting (24%) and underweight (8%). As a result, it had the highest DBM prevalence (27%).
Obesity in children under 5 years of age is still overlooked in the current literature. Our study provided evidence of the increasing burden of obesity in this age group, which was found primarily in households of high socio-economic status. Previous studies reported similar findings arguing that wealthier groups pose strong risk factors for the double burden of malnutrition as well as community-level poverty [48][49][50]. Considering that findings from this study showed that DBM is intertwined in underweight, stunting, wasting and overweight, addressing the social inequalities that share the double burden of child malnutrition in the African region therefore requires strategies that address why certain sub-populations are more exposed to these nutritional problems to avoid strategies that solve one nutritional problem and exacerbate another.
There is a need to increase the engagement of various stakeholders to mitigate the double burden of malnutrition in sub-Saharan Africa. The active collaboration and participation of representatives from local and international non-governmental organizations, major corporations, and government institutions across various sectors such as agriculture, finance, environment, education, communications, health care and nutrition will possibly stimulate dialogues around this menace and proffer solutions and recommendations. Furthermore, progress toward ending hunger and malnutrition by 2030 requires intensified efforts to reduce undernutrition and focused action on reducing obesity and diet-related non-communicable diseases. Key strengths of this study lie in it being a compilation of representative and generalizable DHS datasets from 10 countries. These are typically high-quality, highly responsive datasets from DHS surveys conducted using robust methodologies using well-documented data sources. These DHS surveys are conducted using standardized survey modules and implementations that allow comparisons between countries. However, these are cross-cutting datasets, which limited our ability to assign causality.

Conclusions
In summary, there is a shift in nutrition in Africa, with the increasing prevalence of overweight and obesity among children under five, making optimal child nutrition a key factor in achieving global health goals. The inequality of DBM was consistently pro-poor socio-economic across the ten SSA countries, such that the lower socio-economic groups were more likely to be experiencing DBM and bear a higher burden of the problem than the higher socio-economic groups. In addition, DBM was pro-poor although some of the nutritional indicators were pro-rich. Therefore, the double burden of malnutrition in lowand middle-income countries poses a major global public health problem that could hinder the achievement of the SDGs if not properly addressed. Furthermore, this study has shown the existence of pro-poor inequalities relative to the double burden of malnutrition among under-five children. Therefore, if we are not to leave any child behind, we must address these socio-economic inequalities in sub-Saharan Africa.
Author Contributions: O.A.A. and A.T.L. designed the study, wrote the paper, analyzed data, reviewed the paper, and submitted it for publication. K.N. and E.N. wrote the background section and reviewed all drafts in preparation for publication, D.O. wrote the methods section and reviewed all drafts in preparation for publication, A.S. wrote the discussion section and reviewed all drafts in preparation for publication. P.C. and O.A.S. wrote the paper and reviewed all drafts in preparation for publication. All authors have read and agreed to the published version of the manuscript.
Funding: This research received no external funding; however, the International Journal of Environmental Research and Public Health (IJERPH) supported the study as they waived the APC charges for processing this manuscript.

Institutional Review Board Statement:
This study used secondary analysis based on publicly available DHS datasets. Since the data used in this study were secondary, no ethics approval was sought.

Informed Consent Statement:
The parent study DHS surveys sort the consent from the participants. Data Availability Statement: All data sets are publicly available on the Demographic Health Survey website at: https://dhsprogram.com/what-we-do/survey/survey-display-406.cfm (accessed on 10 January 2023) and can be accessed upon request from the Demographic Health Survey team.