[Skip to Navigation]
Sign In
Figure.  Correlation Between Country-Level Mean Income per Capita and Dietary and Anthropometric Failure (DAF) Category Share
Correlation Between Country-Level Mean Income per Capita and Dietary and Anthropometric Failure (DAF) Category Share

The DAF categories include (A) both failures (BF), (B) dietary failure only (DFO), (C) anthropometric failure only (AFO), and (D) neither failure (NF). The gray shaded areas indicate 95% CI. Income level was measured in US dollars.

Table 1.  Dietary and Anthropometric Failure (DAF) Categoriesa
Dietary and Anthropometric Failure (DAF) Categoriesa
Table 2.  Estimated Prevalence of Children Within DAF Categories Among 162 589 Children Aged 6-23 Months in 55 Countries
Estimated Prevalence of Children Within DAF Categories Among 162 589 Children Aged 6-23 Months in 55 Countries
Table 3.  Share and Estimated Child Population for Each of the DAF Categories Across 55 Countries
Share and Estimated Child Population for Each of the DAF Categories Across 55 Countries
Table 4.  Share of 4 Dietary and Anthropometric Failure (DAF) Types Across Geographic Regions in Europe, Asia, Africa, and South America
Share of 4 Dietary and Anthropometric Failure (DAF) Types Across Geographic Regions in Europe, Asia, Africa, and South America
1.
UNICEF. UNICEF-WHO–The World Bank: joint child malnutrition estimates—levels and trends—2020. March 2020. Accessed August 21, 2020. https://data.unicef.org/resources/jme-report-2020/
2.
Perkins  JM, Kim  R, Krishna  A, McGovern  M, Aguayo  VM, Subramanian  SV.  Understanding the association between stunting and child development in low- and middle-income countries: next steps for research and intervention.   Soc Sci Med. 2017;193:101-109. doi:10.1016/j.socscimed.2017.09.039 PubMedGoogle ScholarCrossref
3.
Patel  PC, Devaraj  S.  Height-income association in developing countries: evidence from 14 countries.   Am J Hum Biol. 2018;30(3):e23093. doi:10.1002/ajhb.23093 PubMedGoogle Scholar
4.
Sudfeld  CR, McCoy  DC, Danaei  G,  et al.  Linear growth and child development in low- and middle-income countries: a meta-analysis.   Pediatrics. 2015;135(5):e1266-e1275. doi:10.1542/peds.2014-3111 PubMedGoogle ScholarCrossref
5.
Horton  S, Steckel  RH.  How Much Have Global Problems Cost the World? A Scorecard from 1900 to 2050. Cambridge University Press; 2013.
6.
Vollmer  S, Harttgen  K, Kupka  R, Subramanian  SV.  Levels and trends of childhood undernutrition by wealth and education according to a Composite Index of Anthropometric Failure: evidence from 146 Demographic and Health Surveys from 39 countries.   BMJ Glob Health. 2017;2(2):e000206. doi:10.1136/bmjgh-2016-000206 PubMedGoogle Scholar
7.
Corsi  DJ, Subramanyam  MA, Subramanian  SV.  Commentary: measuring nutritional status of children.   Int J Epidemiol. 2011;40(4):1030-1036. doi:10.1093/ije/dyr108 PubMedGoogle ScholarCrossref
8.
Ruel  MT, Alderman  H; Maternal and Child Nutrition Study Group.  Nutrition-sensitive interventions and programmes: how can they help to accelerate progress in improving maternal and child nutrition?   Lancet. 2013;382(9891):536-551. doi:10.1016/S0140-6736(13)60843-0 PubMedGoogle ScholarCrossref
9.
Mukhopadhyay  DK, Biswas  AB.  Food security and anthropometric failure among tribal children in Bankura, West Bengal.   Indian Pediatr. 2011;48(4):311-314. doi:10.1007/s13312-011-0057-2 PubMedGoogle ScholarCrossref
10.
Perkins  JM, Subramanian  SV, Davey Smith  G, Özaltin  E.  Adult height, nutrition, and population health.   Nutr Rev. 2016;74(3):149-165. doi:10.1093/nutrit/nuv105 PubMedGoogle ScholarCrossref
11.
Mann  J, Truswell  AS.  Essentials of Human Nutrition. 5th ed. Oxford University Press; 2017.
12.
Perkins  JM, Jayatissa  R, Subramanian  SV.  Dietary diversity and anthropometric status and failure among infants and young children in Sri Lanka.   Nutrition. 2018;55-56:76-83. doi:10.1016/j.nut.2018.03.049 PubMedGoogle ScholarCrossref
13.
Beckerman-Hsu  JP, Chatterjee  P, Kim  R, Sharma  S, Subramanian  SV.  A typology of dietary and anthropometric measures of nutritional need among children across districts and parliamentary constituencies in India, 2016.   J Glob Health. 2020;10(2):020424. doi:10.7189/jogh.10.020424 PubMedGoogle Scholar
14.
Joe  W, Rajpal  S, Kim  R,  et al.  Association between anthropometric-based and food-based nutritional failure among children in India, 2015.   Matern Child Nutr. 2019;15(4):e12830. doi:10.1111/mcn.12830 PubMedGoogle Scholar
15.
Jones  AD, Ickes  SB, Smith  LE,  et al.  World Health Organization infant and young child feeding indicators and their associations with child anthropometry: a synthesis of recent findings.   Matern Child Nutr. 2014;10(1):1-17. doi:10.1111/mcn.12070 PubMedGoogle ScholarCrossref
16.
United Nations. United Nations world population prospects 2019. Accessed September 1, 2020. https://population.un.org/wpp/Download/Standard/Population/
17.
World Health Organization.  Global Nutrition Monitoring Framework: Operational Guidance for Tracking Progress in Meeting Targets for 2025. World Health Organization; 2017.
18.
Croft  TN, Marshall  AM, Allen  CK, Arnold  F, Assaf  S, Balian  S.  Guide to DHS Statistics. ICF; 2018.
19.
Moursi  MM, Arimond  M, Dewey  KG, Trèche  S, Ruel  MT, Delpeuch  F.  Dietary diversity is a good predictor of the micronutrient density of the diet of 6- to 23-month-old children in Madagascar.   J Nutr. 2008;138(12):2448-2453. doi:10.3945/jn.108.093971 PubMedGoogle ScholarCrossref
20.
Wondafrash  M, Huybregts  L, Lachat  C, Bouckaert  KP, Kolsteren  P.  Dietary diversity predicts dietary quality regardless of season in 6-12-month-old infants in south-west Ethiopia.   Public Health Nutr. 2016;19(14):2485-2494. doi:10.1017/S1368980016000525 PubMedGoogle ScholarCrossref
21.
Steyn  NP, Nel  JH, Nantel  G, Kennedy  G, Labadarios  D.  Food variety and dietary diversity scores in children: are they good indicators of dietary adequacy?   Public Health Nutr. 2006;9(5):644-650. doi:10.1079/PHN2005912 PubMedGoogle ScholarCrossref
22.
WHO, UNICEF, USAID, AED, UCDAVIS, IFPRI.  Indicators for Assessing Infant and Young Child Feeding Practices: Part 1: Definitions. World Health Organization; 2008.
Limit 200 characters
Limit 25 characters
Conflicts of Interest Disclosure

Identify all potential conflicts of interest that might be relevant to your comment.

Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.

Err on the side of full disclosure.

If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.

Not all submitted comments are published. Please see our commenting policy for details.

Limit 140 characters
Limit 3600 characters or approximately 600 words
    Original Investigation
    Global Health
    August 12, 2021

    Assessment of Undernutrition Among Children in 55 Low- and Middle-Income Countries Using Dietary and Anthropometric Measures

    Author Affiliations
    • 1Department of Development Economics, Centre for Modern Indian Studies, University of Goettingen, Göttingen, Germany
    • 2Division of Health Policy and Management, Korea University College of Health Science, Seoul, South Korea
    • 3Interdisciplinary Program in Precision Public Health, Department of Public Health Sciences, Graduate School of Korea University, Seoul, South Korea
    • 4Harvard Center for Population and Development Studies, Cambridge, Massachusetts
    • 5Department of Social and Behavioral Sciences, Harvard T.H. Chan School of Public Health, Boston, Massachusetts
    JAMA Netw Open. 2021;4(8):e2120627. doi:10.1001/jamanetworkopen.2021.20627
    Key Points

    Question  Is anthropometric failure an adequate stand-alone measure to estimate global child undernutrition, and how do estimates change when dietary measures are also taken into account?

    Findings  In this cross-sectional study of 162 589 children (aged 6-23 months) in 55 low- and middle-income countries, dietary and anthropometric measures were discordant for 51% of children. A total of 43% of children (equivalent to 45.3 million children) had dietary failure without showing any signs of anthropometric failures.

    Meaning  The findings of this study suggest that the current standard of measuring child undernutrition with anthropometric failure should be complemented with diet- and food-based measures.

    Abstract

    Importance  Evidence on the suitability of anthropometric failure (ie, stunting, underweight, and wasting) as a stand-alone measure of child undernutrition can inform global and national nutrition and health agendas.

    Objective  To provide a comprehensive estimate of the prevalence of child undernutrition by evaluating both dietary and anthropometric measures simultaneously across 55 low- and middle-income countries.

    Design, Setting, and Participants  This was a cross-sectional study that used Demographic and Health Surveys program data from July 2009 to January 2019, to allocate children into dietary and anthropometric failure categories. Nationally representative household surveys were conducted in 55 low- and middle-income countries. Participants included children aged 6 to 23 months who were born singleton and had valid anthropometric measures as well as available 24-hour food intake recollection. Data analysis was conducted from August 23 to October 22, 2020.

    Exposures  Two factors were considered to allocate children into the respective categories. Dietary failure was based on the World Health Organization standards for minimum dietary diversity. Anthropometric failure was constructed using the World Health Organization child growth reference standard z score for stunted growth, muscle wasting, and less than average weight for age.

    Main Outcomes and Measures  Dietary and anthropometric failures were cross-tabulated, which yielded 4 potential outcomes: dietary failure only, anthropometric failure only, both failures, and neither failure. Total child populations for each category were extrapolated from United Nations population estimates.

    Results  Of the 162 589 children (median age [range], 14 months [6-23 months]; 83 467 boys [51.3%]; 78 894 Asian children [48.5%]) in our sample, 42.9% of children had dietary failure according to the standard World Health Organization definition without being identified as having anthropometric failures. In all, 34.7% had both failures, 42.9% had dietary failure only, 8.3% had anthropometric failure only, and 14.1% had neither failure. Dietary and anthropometric measures were discordant for 51.2% of children; these children had nutritional needs identified by only 1 of the 2 measures. Dietary failure doubled the proportion of children in need of dietary interventions compared with anthropometry alone (43%). A total of 45.3 million additional children who experienced undernutrition in these 55 countries were not captured through the evaluation of anthropometric failures only. These results were consistent across geographic regions.

    Conclusions and Relevance  The results of this cross-sectional study suggest that the current standard of measuring child undernutrition by estimating the prevalence of anthropometric failure should be complemented with dietary and food-based measures. Anthropometry alone may fail to identify many children who have insufficient dietary intake.

    Introduction

    Child undernutrition is a significant burden across the globe, with World Health Organization (WHO) estimates reporting more than 205 million children who are undernourished, particularly in low- and middle-income countries (LMICs).1 Undernutrition in early childhood has been linked to significant harm in physical as well as cognitive development.2-5 As such, preventing and treating child undernutrition is not only relevant to achieve the Sustainable Development Goal No. 3 “Good Health and Well-being,” which is part of the United Nations’ 2030 Agenda for Sustainable Development, but also to address root causes of health inequality.

    There are 2 relevant measures regarding the evaluation of children’s nutrition: anthropometry and diet. Although both measures are equally relevant, scientific research and policy agenda most often rely on anthropometry (the study of measurements or proportions of the human body), more specifically, anthropometric failures, when assessing the degree and magnitude of undernourishment among children.6,7 A child is considered to have anthropometric failure if they have stunted growth, are underweight, have muscle wasting, or have a combination of these 3 descriptions. Anthropometric failure is an important measure that is closely related to food and often leads to targeted nutrition-based interventions.8,9 At the same time, it is a fairly complex indicator capturing genetic, environmental and household factors as well.10 This means that anthropometric failure may occur even when nutritional intake is generally sufficient. Similarly, not all nutritional deficiencies would be expected to result in anthropometric failure.11 Thus, although anthropometric failure may indicate undernutrition, it is an imprecise measure for assessing the true extent of undernutrition burden and for identifying precise target groups for nutrition interventions.12-15

    To examine the association between the prevalence of diet and anthropometric failure, Beckerman-Hsu et al13 recently established a typology framework consisting of 4 dietary and anthropometric failure (DAF) categories: dietary failure only (DFO), anthropometric failure only (AFO), both failures (BF), and neither failure (NF) (Table 1). Using Indian Demographic Health Survey (DHS) program data to assign children into these 4 categories, they found that 36.3% of children had micronutrient deficiency without showing any sign of anthropometric failure. Including those children who also showed signs of anthropometric failure, more than 80.3% of children did not meet the WHO standard for minimum acceptable diet (compared with 53.8% who showed signs of anthropometric failure). By using this newly proposed typology, that study13 showed that considering anthropometric measures only to estimate the extent of undernutrition does not capture the full burden and that many children with insufficient micronutrient intake remain “hidden,” which leads to imprecise target groups for nutrition interventions. Our analyses for India showed very similar results (Table 1). Those findings also have important implications for policymakers, for example, when allocating budgets for targeted interventions.

    Given that the study by Beckerman-Hsu et al13 was restricted to India, there is a need to apply the DAF framework to a larger cross-country sample. If the results presented for India can be validated for a global sample, it becomes clear that measures of diet and micronutrient intake need to be much more prevalent in future research in childhood undernutrition. Applying the DAF globally has 2 potential benefits: (1) the identification of children with nutritional need who were previously undetected, giving more precise target groups and estimates of the true extent of child undernutrition, and (2) the potential identification of the precise food-based needs (ie, which micronutrients are missing) that would enable policymakers to make evidence-based prioritization for resource allocation and to monitor the progress of respective interventions at both national and global levels.

    In this study, we extended the DAF framework introduced by Beckerman-Hsu et al13 to 55 LMICs for which DHS data were available. We calculated country-specific typology patterns and derived an estimate of the magnitude of the true burden caused by anthropometric failure and micronutrient deficiency together among children aged 6 to 23 months. We also estimated how many children may be overlooked if the global health community continues to assess child undernutrition based on anthropometric failure alone.

    Methods
    Data

    The DHS program conducts survey waves on population, health, and nutrition for nationally representative samples across the globe. This cross-sectional analysis was reviewed by the Harvard T.H. Chan School of Public Health Institutional Review Board and was considered exempt from full review because the study was based on an anonymous public use data set with no identifiable information on the survey participants. Households were chosen in a 2-stage process that included a selection of enumeration areas and then a sample of households for each enumeration area. Verbal informed consent was sought from respondents by reading a prescribed statement to the respondent and recording in the questionnaire whether the respondent consented. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline for cross-sectional studies. For each country, we used data from the most recent DHS survey wave, which was conducted after July 2009, to January 2019, that contained information on children’s dietary intake and height and weight measurements as well as month-specific age information. These surveys were available for 55 countries. For Colombia, we used the DHS wave 6 data from 2010 instead of wave 7 data from 2015, as it contained many more data points on nutrition relevant to our analyses. We also used population data drawn from the United Nations World Population Prospect (2018)16 to estimate the total child population count corresponding to each DAF type. As such, the 7 largest countries that considered children aged 6 to 23 months (ie, India, Nigeria, Pakistan, Democratic Republic of Congo, Bangladesh, Ethiopia, and Egypt) contributed approximately  66% of the total sample of 55 countries. Mean per capita income information was drawn from PovCalNet (an online analysis tool for monitoring global rates of resource-constrained regions), the United Nations University World Institute for Development Economics Research, and Human Development Reports.

    Study Population and Sample Size

    Our analyses included children aged 6 to 23 months, which is in line with the WHO Indicators for assessing infant and young feeding practices.17 We obtained nationally representative DHS data from 55 LMICs for 208 044 children aged 6 to 23 months who were the youngest child in their household and who lived together with their mother.18 Of our sample of 208 020 children, 166 929 had height and weight measurements used to derive the z scores to identify anthropometric failure. Of these children, 162 589 had dietary data, yielding our final sample size of 162 589 children. Of the final sample size, 83 467 (51.3%) were boys, 79 122 (48.7%) were girls, 56 880 (35.0%) were between 6 and 11 months old, and 105 708 (65.0%) were between 12 and 23 months old (eFigure 1 in the Supplement).

    Outcomes
    Anthropometric Failure

    Children’s anthropometry data (measured height and weight) were obtained from the DHS data. The WHO child growth reference standard z score was used to identify children with stunted growth (height-for-age z score <2), who were underweight (weight-for-age z score <2), and had muscle wasting (weight-for-height z score <2). Anthropometric failure was defined as a binary variable with anthropometric failure prevailing if a child had stunted growth, was underweight, had muscle wasting, or a combination of these failures.

    Dietary Failure

    Children’s dietary data were based on a 24-hour recall in the DHS surveys. Children’s consumption of the following 8 food groups was collected: (1) grains, roots, and tubers; (2) legumes and nuts; (3) dairy products (milk, yogurt, cheese); (4) flesh foods (meat, fish, poultry, and liver/organ meats); (5) eggs; (6) vitamin A–rich fruits; (7) vegetables; and (8) breast milk. We defined dietary failure as a binary variable, assigning the outcome yes if the child’s dietary intake did not meet the minimum dietary requirements as defined by the WHO classification from 2017, which required an intake of a minimum of 5 of 8 different food categories.17 As recommended by the DHS,18 food categories with answers such as “do not know” or that were missing individual data points were assigned an outcome of no. Given our sample of approximately 200 000 responses for each food category and approximately 350 responses for each category being either do not know or missing data points, we do not believe that assigning the missing 0.2% to the no category significantly changed our results. The indicator for minimum dietary diversity was created as a way to use dietary data to capture the micronutrient density of the diets in children aged 6 to 23 months and has been validated previously.19,20 Therefore, children who did not reach minimum dietary diversity were considered to have unmet nutritional need regardless of prevailing anthropometric failure.

    Statistical Analysis

    We cross-tabulated anthropometric and dietary failures yielding 4 potential outcomes: DFO, AFO, BF, and NF. We calculated the prevalence of DAF. For each respective DAF category, we calculated the prevalence of DAF on national levels for all 55 LMICs. Additionally, we estimated the burden, in terms of total number of children, of each DAF category by using United Nations population data from 2018 for children aged 0 to 5 years for these 55 countries. Given that DHS is a nationally representative data set, we extrapolated the share of children aged 6 to 23 months (30.3%) from children aged 0 to 5 years (353.0 million) to the national population estimates according to the United Nations, yielding our final population size (105.8 million) of children aged 6 to 23 months for the 55 countries included in our analysis. We considered 5% levels of significance, and hypothesis tests were 2-sided. All analyses were conducted on Stata, version 16.0 (StataCorp). These data were analyzed from August 23 to October 22, 2020.

    Results
    Overall Child Undernutrition Using DAF Categories

    This study included a total of 162 589 children (median age [range], 14 months [6-23 months]; 83 467 boys [51.3%]; 78 894 Asian children [48.5%]). Across all 55 LMICs and weighted by country size, 77.6% of the total sample of children aged 6 to 23 months were shown to have dietary failure, whereas only 43.0% had at least 1 form of anthropometric failure. The most common category was DFO in 42.9% of children, followed by BF in 34.7%, NF in 14.1%, and AFO in 8.3% (Table 2). Dietary and anthropometric measures were discordant for 51.2% of children; these children had nutritional needs identified by only 1 of the 2 measures (DFO + AFO). Although these results were strongly influenced by India, which accounted for roughly one-third of the total child population in our final sample, the results did not change significantly for DFO when we considered unweighted averages (DFO, 45.9%; BF, 26.4%; NF, 19.7%; AFO, 8.1%) (eTable 1 in the Supplement).

    Using the weighted prevalence of DAF in Table 2, we estimated the total number of children in different categories of DAF and found that DFO was the largest category, with an estimated population of 45.3 million children, followed by BF with 36.7 million children, NF with 14.9 million children, and AFO with 8.8 million children (Table 2). A more granular look at the prevalence of dietary failure and the prevalence of individual causes of anthropometric failure (ie, stunted growth, muscle wasting, and lower-than-average weight) can be found in eTable 2 in the Supplement.

    Country-Specific Analysis of DAF

    Table 3 shows country-specific estimates for the prevalence of the DAF categories together with estimated population numbers. A certain degree of variation in the share of each DAF category was found for different countries. Although the Maldives (51.5%) and Peru (57.2%) had fairly large shares of children with NF, Niger (53.6%), Burkina Faso (47.9%), and Burundi (47.6%) had BF for roughly half of their child population aged 6 to 23 months. Gabon (66.2%), Haiti (62.6%), and Liberia (60.9%) had large shares of children in the DFO category, which captures children with nutritional needs who are missed when only anthropometric measures are evaluated. For 41 of the 55 LMICs, DFO was the largest category. For 38 LMICs, at least 40% of children were in the DFO category, further highlighting the importance of capturing nutritional intake in addition to anthropometric failure. In terms of the total number of children with DFO, the largest 8 countries contributed approximately 66.7% to the total number of children with DFO. Of those, India was the largest, contributing 28.0% to the total DFO number (Table 3).

    DAF by Geographic Region and Country-Level Income

    The level of a country’s share of children with DFO appeared to be consistent across different geographic regions and income levels. Although NF and BF shares varied widely, DFO accounted for most children in all geographic regions, ranging from approximately  35% of children in Central and South America to approximately  50% in Europe (ie, Albania and Armenia) (Table 4). Country-level income levels (measured as mean annual household income per capita) did not appear to substantially affect the share of children with DFO (Figure). Although the prevalences of BF and NF were significantly associated with a country’s income level, DFO was associated at a 10% level of significance, and AFO had no significant association. The distribution of DAF category shares for different mean income per capita levels is displayed in eFigure 2 in the Supplement.

    Variation Within DAF

    It may take time for anthropometric failure to manifest in children; thus, children with BF may typically be older than children with DFO. At the same time, certain food groups may explain the allocation into the respective DAF categories, such that certain food groups may be responsible for DFO, AFO, and BF. Therefore we evaluated these 2 factors as a part of the robustness testing for this study.

    eTable 3 in the Supplement shows the average age in months for each of the DAF categories across the region. The category AFO had the oldest average age in months across all regions, and DFO had the lowest average age in 3 of the 4 regions. However, when we compared the differences between DFO and BF (because both have dietary failure and thus are a better comparison than DFO and AFO), the variation was moderate: between −0.1 to 1.7 months of average age.

    eFigures 3 and 4 in the Supplement present the variation of average consumption level of the previously mentioned 8 different food groups, both for the whole sample and for each region. Average consumption patterns for each of the food groups were similar for both categories without dietary failure (AFO and NF) as well as the categories with dietary failure (DFO and BF). This finding suggests that there was not a particular micronutrient that was responsible for the anthropometric failure. Average consumption levels were particularly low for legumes and nuts, flesh foods, eggs, and vegetables. Overall, although certain groups were more responsible for causing dietary failure, there was no indication that certain micronutrients cause anthropometric failure, which suggests that micronutrient intake should be examined next to anthropometric failure to capture nutrition deficiencies.

    Discussion

    Our cross-sectional study found that approximately 4 of every 10 children in our analytic sample of 55 LMICs had no anthropometric failures but were identified as having dietary failure (the DFO category); this result included more than 45 million children aged 6 to 23 months with unmet dietary need who were not identified by measures focused solely on anthropometry. Approximately one-third of children had both dietary and anthropometric failures, and 4 of 5 children did not meet the minimum dietary diversity as recommended by WHO, which suggests the depth of nutritional need among a large proportion of the global child population in LMICs. Although there was some variation across countries, the relevance of the DFO category was consistent across different geographic regions and country-level mean income, as was within-category variation of age and food group consumption.

    Future research will be needed to identify improved ways of measuring dietary intake, ideally over a longer period of time to account for the fact that the past 24 hours may not be reflective of regular daily alimentation. We used the WHO minimum dietary diversity indicator, but other dietary indicators may be considered, particularly those that combine the cost-effectiveness of a survey question (rather than relying on biomedical information and medical examination results) with increased reliability over time and sensitivity.21 Similarly, given varying degrees of nutritional intake at different stages in early childhood, other indicators may be considered for different age groups. Given that Beckerman-Hsu et al13 found larger variations of the DAF categories at the district level in India, future research must also take a more granular look into county- and district-level DAF prevalence to properly choose and prioritize food policy interventions.

    The fact that DFO was found to be a primary category across geographic regions and country-level income suggests the need to consider dietary intake along with anthropometric failure when assessing the extent of nutritional burden in a more comprehensive manner and to successfully foster food security across the globe. We acknowledge that anthropometric failure is an important measure that is associated with nutrition and often leads to targeted interventions that include providing food to the children in need.8,9 We therefore do not intend to argue for a replacement or substitution of anthropometric failure in global health research and policy. Instead, our study findings suggest that the consideration of anthropometric failures alone may leave undetected large parts of the population with nutritional deficiencies. Across the 55 LMICs considered in our analysis, this translated to more than 45 million children who appeared to not have a nutritional deficiency based on anthropometry measures but in fact were in need of better nutrition. Analysis of both dietary intake and anthropometric failure may enable more precise identification of children with nutritional needs and may facilitate policymakers to develop more effective, equitable, and targeted interventions that could increase global food security.

    Limitations

    Our study had 2 sources of data limitations. First, given that the dietary data were self-reported by mothers based on a 24-hour recall, the innate nature of the data was subject to some measurement error. However, DHS data on dietary intake have been found to be appropriate for the population level.22 Second, our estimates may have been biased by survey nonresponse and missing data for specific survey items or countries. However, given that we obtained complete nutrition and anthropometric data for approximately 80% of all children in the sample for 55 of the 60 countries that conducted the standard DHS surveys in the past 10 years (data from Afghanistan, Philippines, Jordan, Indonesia, and Madagascar were missing), any bias is expected to be small.

    Conclusions

    The results of this cross-sectional study suggest that the current standard of measuring child undernutrition should include diet- and food-based measures because anthropometry alone failed to identify many children who have insufficient dietary intake. The findings further suggest that, like anthropometric failure, dietary diversity and micronutrient intake should be considered when assessing the global nutritional status of children.

    Back to top
    Article Information

    Accepted for Publication: June 7, 2021.

    Published: August 12, 2021. doi:10.1001/jamanetworkopen.2021.20627

    Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2021 Heemann M et al. JAMA Network Open.

    Corresponding Author: Sebastian Vollmer, PhD, Department of Development Economics, Centre for Modern Indian Studies, University of Goettingen, 26 Waldweg, 37073 Göttingen, Germany (sebastian.vollmer@wiwi.uni-goettingen.de); S. V. Subramanian, PhD, Harvard Center for Population and Development Studies, Nine Bow St, Cambridge, MA 02138 (svsubram@hsph.harvard.edu).

    Author Contributions: Mr Heemann and Dr Kim had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

    Concept and design: All authors.

    Acquisition, analysis, or interpretation of data: All authors.

    Drafting of the manuscript: Heemann.

    Critical revision of the manuscript for important intellectual content: Kim, Vollmer, Subramanian.

    Statistical analysis: Heemann, Kim.

    Obtained funding: Heemann, Subramanian.

    Supervision: Vollmer, Subramanian.

    Conflict of Interest Disclosures: None reported.

    Funding/Support: This study was supported in part by the Open Access Publication Funds of the University of Goettingen.

    Role of the Funder/Sponsor: The University of Goettingen had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

    Additional Information: We acknowledge the support of the Demographic and Health Surveys Program for providing access to the data.

    References
    1.
    UNICEF. UNICEF-WHO–The World Bank: joint child malnutrition estimates—levels and trends—2020. March 2020. Accessed August 21, 2020. https://data.unicef.org/resources/jme-report-2020/
    2.
    Perkins  JM, Kim  R, Krishna  A, McGovern  M, Aguayo  VM, Subramanian  SV.  Understanding the association between stunting and child development in low- and middle-income countries: next steps for research and intervention.   Soc Sci Med. 2017;193:101-109. doi:10.1016/j.socscimed.2017.09.039 PubMedGoogle ScholarCrossref
    3.
    Patel  PC, Devaraj  S.  Height-income association in developing countries: evidence from 14 countries.   Am J Hum Biol. 2018;30(3):e23093. doi:10.1002/ajhb.23093 PubMedGoogle Scholar
    4.
    Sudfeld  CR, McCoy  DC, Danaei  G,  et al.  Linear growth and child development in low- and middle-income countries: a meta-analysis.   Pediatrics. 2015;135(5):e1266-e1275. doi:10.1542/peds.2014-3111 PubMedGoogle ScholarCrossref
    5.
    Horton  S, Steckel  RH.  How Much Have Global Problems Cost the World? A Scorecard from 1900 to 2050. Cambridge University Press; 2013.
    6.
    Vollmer  S, Harttgen  K, Kupka  R, Subramanian  SV.  Levels and trends of childhood undernutrition by wealth and education according to a Composite Index of Anthropometric Failure: evidence from 146 Demographic and Health Surveys from 39 countries.   BMJ Glob Health. 2017;2(2):e000206. doi:10.1136/bmjgh-2016-000206 PubMedGoogle Scholar
    7.
    Corsi  DJ, Subramanyam  MA, Subramanian  SV.  Commentary: measuring nutritional status of children.   Int J Epidemiol. 2011;40(4):1030-1036. doi:10.1093/ije/dyr108 PubMedGoogle ScholarCrossref
    8.
    Ruel  MT, Alderman  H; Maternal and Child Nutrition Study Group.  Nutrition-sensitive interventions and programmes: how can they help to accelerate progress in improving maternal and child nutrition?   Lancet. 2013;382(9891):536-551. doi:10.1016/S0140-6736(13)60843-0 PubMedGoogle ScholarCrossref
    9.
    Mukhopadhyay  DK, Biswas  AB.  Food security and anthropometric failure among tribal children in Bankura, West Bengal.   Indian Pediatr. 2011;48(4):311-314. doi:10.1007/s13312-011-0057-2 PubMedGoogle ScholarCrossref
    10.
    Perkins  JM, Subramanian  SV, Davey Smith  G, Özaltin  E.  Adult height, nutrition, and population health.   Nutr Rev. 2016;74(3):149-165. doi:10.1093/nutrit/nuv105 PubMedGoogle ScholarCrossref
    11.
    Mann  J, Truswell  AS.  Essentials of Human Nutrition. 5th ed. Oxford University Press; 2017.
    12.
    Perkins  JM, Jayatissa  R, Subramanian  SV.  Dietary diversity and anthropometric status and failure among infants and young children in Sri Lanka.   Nutrition. 2018;55-56:76-83. doi:10.1016/j.nut.2018.03.049 PubMedGoogle ScholarCrossref
    13.
    Beckerman-Hsu  JP, Chatterjee  P, Kim  R, Sharma  S, Subramanian  SV.  A typology of dietary and anthropometric measures of nutritional need among children across districts and parliamentary constituencies in India, 2016.   J Glob Health. 2020;10(2):020424. doi:10.7189/jogh.10.020424 PubMedGoogle Scholar
    14.
    Joe  W, Rajpal  S, Kim  R,  et al.  Association between anthropometric-based and food-based nutritional failure among children in India, 2015.   Matern Child Nutr. 2019;15(4):e12830. doi:10.1111/mcn.12830 PubMedGoogle Scholar
    15.
    Jones  AD, Ickes  SB, Smith  LE,  et al.  World Health Organization infant and young child feeding indicators and their associations with child anthropometry: a synthesis of recent findings.   Matern Child Nutr. 2014;10(1):1-17. doi:10.1111/mcn.12070 PubMedGoogle ScholarCrossref
    16.
    United Nations. United Nations world population prospects 2019. Accessed September 1, 2020. https://population.un.org/wpp/Download/Standard/Population/
    17.
    World Health Organization.  Global Nutrition Monitoring Framework: Operational Guidance for Tracking Progress in Meeting Targets for 2025. World Health Organization; 2017.
    18.
    Croft  TN, Marshall  AM, Allen  CK, Arnold  F, Assaf  S, Balian  S.  Guide to DHS Statistics. ICF; 2018.
    19.
    Moursi  MM, Arimond  M, Dewey  KG, Trèche  S, Ruel  MT, Delpeuch  F.  Dietary diversity is a good predictor of the micronutrient density of the diet of 6- to 23-month-old children in Madagascar.   J Nutr. 2008;138(12):2448-2453. doi:10.3945/jn.108.093971 PubMedGoogle ScholarCrossref
    20.
    Wondafrash  M, Huybregts  L, Lachat  C, Bouckaert  KP, Kolsteren  P.  Dietary diversity predicts dietary quality regardless of season in 6-12-month-old infants in south-west Ethiopia.   Public Health Nutr. 2016;19(14):2485-2494. doi:10.1017/S1368980016000525 PubMedGoogle ScholarCrossref
    21.
    Steyn  NP, Nel  JH, Nantel  G, Kennedy  G, Labadarios  D.  Food variety and dietary diversity scores in children: are they good indicators of dietary adequacy?   Public Health Nutr. 2006;9(5):644-650. doi:10.1079/PHN2005912 PubMedGoogle ScholarCrossref
    22.
    WHO, UNICEF, USAID, AED, UCDAVIS, IFPRI.  Indicators for Assessing Infant and Young Child Feeding Practices: Part 1: Definitions. World Health Organization; 2008.
    ×