key: cord-0288601-zhmdxud4
authors: Caleyachetty, R.; Kumar, N.; Bekele, H.; Manaseki-Holland, S.
title: Socioeconomic inequality in the double burden of child malnutrition in the Eastern and Southern African Region
date: 2022-04-02
journal: nan
DOI: 10.1101/2022.03.30.22273164
sha: e7b865e37fa91b00b21c9e73377bae4da7da163d
doc_id: 288601
cord_uid: zhmdxud4

Socioeconomic inequalities in the double burden of child malnutrition threatens global nutrition targets 2025, especially in Eastern and Southern Africa. We aimed to quantify these inequalities from nationally representative household surveys in 13 Eastern and Southern African countries between 2000 and 2018. 13 of the latest Demographic and Health Surveys including 72,231 children under five year olds were studied. Prevalence of stunting, wasting and overweight (including obesity) were disagregated by wealth quintiles, maternal education categories and urban-rural residence for visual inspection of inequalities, and the slope index of inequality (SII) and the relative index of inequality (RII) were estimated for each country. Country-specific estimates were pooled using random-effects meta-analyses. Regional stunting and wasting prevalence was higher among children living in the poorest households, with mother's with the lowest educational level and in rural areas. In contrast, regional overweight (including obesity) prevalence was higher among children living in the richest households, with mother's with the highest educational level and urban areas. Tackling social inequalities in the distribution of the double burden of malnutrition among children in the Eastern and Southern African region will require strategies that address the reasons socially disadvantaged children became more exposed to stunting or wasting.

49 Introduction 50 The African region has the highest burden of childhood stunting and one of the highest . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 2, 2022. ; https://doi.org/10.1101/2022.03.30.22273164 doi: medRxiv preprint 51 burdens of childhood overweight in Africa. 1 Importantly, the region is off track to achieve 52 the UN Sustainable Development Goal (SDG) 2 aim to end all forms of hunger and 53 malnutrition by 2030. 2 The double burden of malnutrition, is considered a major global 54 health challenge for African countries, 3 particularly those in Eastern and Southern Africa.

55 The prevalence of malnutrition is 34.5% in Eastern Africa for stunting and 10.2% in 56 Southern Africa for overweight (including obesity), compared to 29.1% for stunting and 61 for the African region is tackling social inequalities in health. 4 Considering that the double 62 burden of malnutrition is inextricably bound to socioeconomic conditions, 5 monitoring 63 and description of child malnutrition inequalities will be important to serve 64 evidence-based program and policy decisions in the Eastern and Southern African 65 region. Several measures of social inequality in health reflecting its different dimensions 66 have been described. 6 In the context of child malnutrition, absolute inequality measures 67 reflects the magnitude of difference in malnutrition prevalence between two subgroups 68 and relative inequality measures show proportional differences in child malnutrition 69 prevalence among subgroups. The choice of measure is not value-neutral and 70 influences our understanding of which populations may be actually experiencing a higher 71 malnutrition burden. Therefore, both relative and absolute measures to monitor social 72 inequalities are recommended. 7 73 A recent study examined global inequalities in household-level double burden of 74 malnutrition and found the probability of double burden of malnutrition higher among 75 richer households in poorer low-income and middle-income countries. 8 Their definition of . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted April 2, 2022. ; https://doi.org/10.1101/2022.03.30.22273164 doi: medRxiv preprint 76 double burden of malnutrition was expressed as a stunted child with an overweight 77 mother. The objective of reducing child malnutrition may not necessarily be compatible 78 with the objective of reducing socioeconomic inequalities in child malnutrition. For 79 example, Africa has observed spectacular gains in reducing stunting, however these 80 gains may have not benefited every child equally. 9 In the Eastern and Southern African 81 region, there has not been to date a comprehensive large-scale assessment of 82 socioeconomic inequalities in the population-level double burden of child malnutrition.

83 Such an assessment would be valuable to support countries moving towards regional 84 nutrition equity. 

. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted April 2, 2022. 133 134 135 Statistical analysis 136 Regional prevalence estimates and 95% confidence intervals (CIs) for stunting, wasting, 137 overweight (including obesity), concurrent stunting and wasting, and concurrent stunting 138 and overweight (including obesity) in children <5 years were calculated. First, we 139 calculated country-specific malnutrition estimates. The DHS used complex sample 140 designs therefore we accounted for stratification and clustering. Meta-analysis of . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 2, 2022. ; https://doi.org/10.1101/2022.03.30.22273164 doi: medRxiv preprint 141 country-specific estimates were performed to calculate the regional prevalence 142 estimates and 95% CIs. Variances of the raw proportions were stabilized with a double 143 arcsine transformation and then pooled on the basis of a random-effects model.

145 Social inequalities in the population-level double burden of child malnutrition were first 146 visualized using equiplots. Equiplots allow comparison of absolute social inequality 147 indicating both the prevalence of malnutrition in each group and the distance between 148 groups, which represents the absolute social inequality. We then estimated using 149 absolute and relative differences between poorest and richest households, lowest and 150 highest maternal education and urban and rural residence. Wealth and education 151 inequalities in the double burden of malnutrition were measured for each country by 152 regression-based inequality measures, the relative index of inequality (RII) and the slope 153 index of inequality (SII), respectively. 11 A logistic regression model was used to assess 154 the association between malnutrition and wealth or maternal education and to generate 155 the SII and RII values and 95% CIs. Social inequalities on both scales were reported 156 because conclusions can be skewed when only one or the other is used. 12 Both these 157 measures take into account the size of the population across wealth and education 158 groups. The SII can be interpreted as the defined as the absolute difference in the 159 malnutrition prevalence between the poorest and richest or mothers with the lowest 160 education and mothers with the highest education level (across the entire socioeconomic 161 or maternal education distribution). SII can range from -1 to 1. If there is no inequality, 162 SII takes the value zero. Positive SII values indicate that the child malnutrition indicator 163 is concentrated in the most socially advantaged group (i.e. richest or mothers with the 164 highest education). The RII is defined as the ratio of the estimated prevalence of 165 malnutrition between the poorest and the richest or mothers with the lowest education 166 and mothers with the highest education (across the entire socioeconomic or maternal . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 2, 2022. 205 (Namibia) to 39.3% (Mozambique) (S1 Table) ; wasting prevalence was 4.6% (95% CI: 206 3.7-5.6), with prevalence ranging from 2.3% (Rwanda) to 12.3% (Comoros) (S2 Table) ; 207 and overweight (including obesity) prevalence was 6.5% (95% CI 5.2,7.9), with 208 prevalence ranging from 3.6% (Tanzania) to 12.9% (South Africa) (S3 Table) . The 209 pooled prevalence of concurrent stunting and wasting was 1.1% (95% CI 0.8,1.4), with 210 prevalence ranging from 0.4% (South Africa) to 2.1% (Comoros) (S4 Table) . The pooled 211 prevalence of concurrent stunting and overweight (including obesity) was 2.6% (95% CI 212 1.9, 3.5), with the prevalence varying across countries, ranging from 1.1% (Kenya) to 213 5.1% (Comoros) (S5 Table) . is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 2, 2022. 218 Eastern and Southern African countries had prevalences of overweight (including 219 obesity) and stunting or wasting greater than the overall pooled prevalence estimates for 220 overweight (including obesity) and stunting or wasting, respectively. 221 222 Fig 3 presents equiplots of regional stunting prevalence by wealth quintiles, categories of 223 formal maternal education, and urban-rural residence. Higher stunting prevalence was 224 found among children belonging to the poorest households, mother's with the lowest 225 educational level and rural residence. Lesotho represents an extreme example with the 226 widest gaps in stunting prevalence by wealth (30.5% points; p<0.001) and maternal 227 education (54.6% points; p<0.001) (S6-S7 Tables). Malawi showed the widest gap in 228 stunting prevalence by urban-rural residence gap (13.7% points; p<0.001)(S8 Table) . 236 Mozambique showed the widest gap in wasting prevalence by urban-rural residence 237 (3.0%; p<0.001) (S11 Table) . is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 2, 2022. 286 households in richer LMICs. 8 The authors defined the double burden of malnutrition at 287 the household level, expressed as a stunted child with an overweight mother. However, 288 there is strong evidence that stunted child/overweight mother pairs represents a 289 statistical artefact and not a distinct entity. 13 Our study assessed socioeconomic 290 inequalities in the double burden of child malnutrition at the population level, which is a 291 common double burden of malnutrition operational definition. 14 A systematic review 292 examining socioeconomic status and overweight or obesity among older children (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19) 293 years) in Sub-Saharan Africa used the RII to assess socioeconomic inequalities. The 294 school-age children from higher socioeconomic households tended to be overweight and 295 obese. 15 We found similar wealth and education inequalities using both simple and 296 complex absolute and relative measures in the Eastern and Southern African region for . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 2, 2022. ; https://doi.org/10.1101/2022.03.30.22273164 doi: medRxiv preprint 297 children under five year olds. In highly developed countries, a higher prevalence of child 298 overweight and obesity children is typically found in lower socioeconomic groups. 16 299 Eastern and Southern African households of lower socioeconomic status, may have 300 financial constraints that prevent purchasing of high-energy density foods with a 301 consequence that the overweight and obesity prevalence levels are kept to a minimum 302 in these households. Additionally, parent's socio-cultural beliefs may influence their 303 child's body shape ideals. In some sub-Saharan African countries, overweight and 304 obesity have been historically been considered to be a sign of wealth. 15 305 306 Our study found the Eastern and Southern African region is burdened by multiple forms 307 of malnutrition. The prevalence of stunting among children was 30.4%, the prevalence of 308 wasting was 4.6%, whereas 6.5% were overweight (including obesity). A much smaller 309 proportion of children also had concurrent stunting and wasting (1.1%) and concurrent 310 stunting and overweight (2.6%). The presence of a double burden of malnutrition at a 311 population level, is likely to reflect the region's ongoing challenges with poverty, food 312 insecurity, infectious diseases, droughts, floods and conflict as well as the presence of 313 the obesogenic environment driven by globalization and rapid urbanization. Households 314 of lower socioeconomic households may not have access to resources (money, 315 knowledge, prestige, power and beneficial connections) that could better able them to 316 avoid undernutrition and therefore unable to avoid child undernutrition.

318 Several limitations, however, should be considered when interpreting our findings. The 319 Eastern and Southern African region encompasses 24 countries, however our findings 320 are for 13 countries of 17 in the Eastern and Southern African region where DHS is 321 available. A direct indicator of household income is not collected in the DHS and the 322 asset-based household wealth index was used. While this is an imperfect measure . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 2, 2022. ; https://doi.org/10.1101/2022.03.30.22273164 doi: medRxiv preprint 323 financial resources, it is frequently used and judged to be superior to income in lower 324 income countries. 17 Between country comparisons may be affected by the period of data 325 collection for each national household survey (from 2003 to 2018). Social reform and 326 safety net programmes in the region have increased over this study period and may 327 have reduced inequalities over time. 18 Despite these limitations, our study has several 328 strengths. Our analysis of socioeconomic inequalities and urban/rural inequalities in child 

338 Addressing social inequalities in the distribution of the double burden of malnutrition 339 among children in the Eastern and Southern African region requires strategies that 340 address the reasons certain subpopulations became more exposed to these nutrition 341 problems, while also avoiding strategies that solve one nutrition problem while 342 worsening another one. 5 Any attempt to analyse and act on inequalities in the double 343 burden of child malnutrition in the Eastern and Southern African region, however 344 requires the initial acknowledgement that tackling overweight (including obesity) among 345 socially advantaged children, is not part of the "health inequalities" agenda, which 346 focuses on the improving the health of socially disadvantaged. 19 Most countries also do 347 not have a robust routine data collection systems for child malnutrition, 20 and therefore 348 cannot adequately and frequently monitor socioeconomic gradients in child malnutrition.

. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted April 2, 2022. ; https://doi.org/10.1101/2022.03.30.22273164 doi: medRxiv preprint

United Nations Children's Fund (UNICEF), World Health Organization, International Bank for Reconstruction and Development/The World Bank. Levels and trends in child malnutrition: key findings of the 2021 edition of the joint child malnutrition estimates. Geneva: World Health Organization

Unlocking the potential for achievement of the UN Sustainable Development Goal 2 -'Zero Hunger' -in Africa: targets, strategies, synergies and challenges

Regional Overview on the Double Burden of Malnutrition and Examples of Program and Policy Responses: African Region

Global action on the social determinants of health

Nutrition disparities and the global burden of malnutrition

Measures of health inequalities: part 2

The social determinants of health: Developing an evidence base for political action. Final report of the Measurement and Evidence Knowledge Network to the World Health Organization Commission on the Social Determinants of Health

Global inequalities in the double burden of malnutrition and associations with globalisation: a multilevel analysis of Demographic and Healthy Surveys from 55 low-income and middle-income countries

Global Nutrition Report: The state of global nutrition

Demographic and Health Surveys

Relative index of inequality and slope index of inequality: a structured regression framework for estimation

Use of relative and absolute effect measures in reporting health inequalities: structured review

Stunted child/overweight mother pairs represent a statistical artifact, not a distinct entity

The Double Burden of Malnutrition: A Systematic Review of Operational Definitions

Socio-economic status and overweight or obesity among school-age children in sub-Saharan Africa -a systematic review

Socioeconomic factors and childhood overweight in Europe: results from the multi-centre IDEFICS study

Realizing the Full Potential of Social Safety Nets in Africa. Africa Development Forum series

Health disparities and health equity: the issue is justice

Review of National Nutrition Surveillance Systems

The COVID-19 crisis will exacerbate maternal and child undernutrition and child mortality in low-and middle-income countries