henhaobio
Sign in

PLoS ONEDevelopmental BiologyPeer-reviewed journal article

Determinants of Minimum Acceptable Diet among 6-23 Months Age Children in Ethiopia: A Multilevel Analysis of The Ethiopian Demographic Health Survey

Aberash Abay Tassew, Dejen YemaneiD, Abate Bekele BelachewiD, Beyene Meressa AdhenaiD

PDFDOIPreprintData
In-depth full-text digest available · Jump to digest ↓

Abstract

BackgroundThough infant and young children should be fed according to a minimum acceptable diet to ensure appropriate growth and development, only 7% of Ethiopian 6-23 months age children meet the minimum acceptable dietary standards, which is lower than the national target of 11% set for 2016. Therefore, this study aims to assess the individual and community level determinants of minimum acceptable diet among 6-23 months age children in Ethiopia.\n\nMethodsThis study analyzed retrospectively a cross-sectional data on a weighted sample of 2919 children aged 6-23 months nested within 617 clusters after extracting from Ethiopian Demographic and Health Survey 2016 via the link www.measuredhs.com. By employing bi-variate multilevel logistic regression model, variables which were significant at the p-value < 25 were included in multivariable multilevel logistic regression analysis. Finally, variables with p-value < 0.05 were considered as significant predictors of minimum acceptable diet.\n\nResultsOnly 6.1% of 6-23 months age children feed minimum acceptable diet in Ethiopia. Children 18-23 months age (AOR=3.7, 95%CI 1.9, 7.2), fathers with secondary or higher education (AOR=2.1, 95%CI 1.2, 3.6), Employed mothers (AOR=1.7, 95%CI 1.2, 2.5), mothers have access to drinking water (AOR=1.9, 95%CI 1.2, 2.9), mothers with media exposure (AOR=2.1 95%CI 1.1, 2.7) were positive individual level predictors. Urban mothers (AOR=4.8, 95%CI 1.7, 13.2)) and agrarian dominant region (AOR=5.6, 95%CI 2.2, 14.5) were community level factors that significantly associated with minimum acceptable diet of 6-23 months age children.\n\nConclusionBoth individual and community level factors were significantly associated with minimum acceptable diet of 6-23 months age children in Ethiopia, suggesting that nutritional interventions designed to improve child health should not only be implemented at the individual level but tailored to community context as well.

In-depth summary

Claude Opus 5.5 · xhigh effort · 8 Nov 2026 · from the full text of version 1 (16 Aug 2018) · CC BY 4.0

Only 6.1% of Ethiopian 6-23-month-olds received a minimum acceptable diet (EDHS 2016); child age, paternal education, maternal employment, water access, media exposure and region were associated with it.

Minimum acceptable diet (MAD), a WHO complementary-feeding indicator combining dietary diversity and meal frequency, rose only from 3% to 7% in Ethiopia over 2005-2016 despite national nutrition programmes. Earlier Ethiopian studies were small, not nationally representative and used single-level logistic regression on clustered data, so this analysis applies two-level mixed-effects logistic regression to the Ethiopian Demographic and Health Survey (EDHS) 2016.

Key results

  1. ShownFew Ethiopian children aged 6-23 months met the WHO minimum acceptable diet in 2016. Many more met the meal-frequency requirement than the dietary-diversity requirement.

    6.1% of 2919 children were fed MAD. 11% met minimum dietary diversity (four or more of seven food groups) and 42% met minimum meal frequency (Table 4). The background cites 7% from the survey report against a national 2016 target of 11%. · Minimum acceptable diet feeding practice; Abstract; Study variables

  2. ShownOlder children, and children whose fathers had secondary or higher education, had higher adjusted odds of receiving MAD.

    In the multilevel analysis (Table 5): children aged 18-23 vs 6-8 months, adjusted odds ratio (AOR) 3.7 (95% CI 1.9-7.3); fathers with secondary or higher education, AOR 2.4 (95% CI 1.4-4.0). The abstract instead gives AOR 3.7 (1.9, 7.2) and AOR 2.1 (1.2, 3.6) for the same factors. · Individual and community level determinants of…; Abstract; Discussion

  3. ShownEach of these was linked to higher odds of MAD: an employed mother, a drinking-water trip under 30 minutes, and maternal media exposure.

    Employed vs unemployed mothers: AOR 1.7 (95% CI 1.2-2.5). Under 30 minutes to fetch drinking water vs longer: AOR 2.0 (1.3-3.1). Media exposure vs none: AOR 1.95 (1.3-2.9). The abstract reports different values for water (AOR 1.9, 1.2-2.9) and media (AOR 2.1, 1.1-2.7). · Individual and community level determinants of…; Abstract; Discussion

  4. SuggestedRegion of residence was the significant community-level predictor. Children in agrarian-dominant regions and in cities had higher odds of MAD than those in pastoralist-dominant regions.

    Agrarian vs pastoralist: AOR 5.1 (95% CI 2.0-13.1). City dwellers vs pastoralist: AOR 5.4 (1.9-14.8). The abstract instead lists urban residence (AOR 4.8, 1.7-13.2) and agrarian region (AOR 5.6, 2.2-14.5). 91.8% of mothers lived in agrarian-dominant regions. · Individual and community level determinants of…; Abstract; Community level characteristics of the study…

  5. ShownCommunities explained a large share of the variation in MAD feeding, which justified the multilevel design. Significant between-community variance remained after adjusting for all covariates.

    Null-model intraclass correlation (ICC) was about 40% (39.6%, P<0.001). Between-community variability fell to 28.3% (individual-only model), 27.5% (community-only) and 26.6% (combined). The combined model had the lowest AIC and a reported 44.9% explained variance. Residual community variance stayed significant (p<0.001). · Statistical analysis; Discussion

Methods, data and reproducibility

Secondary analysis of the cross-sectional EDHS 2016 child recode. A binary, WHO-defined MAD outcome was modelled by two-level random-intercept logistic regression (children within clusters) after bivariable screening. Community variables were cluster averages of mothers' characteristics. Four models were compared by intraclass correlation, proportional change in variance and AIC.

  • EDHS 2016 child recode: children aged 6-23 months living with their mother and with complete MAD data (26 not living with their mother and 98 with incomplete MAD data excluded) · human · n = 2919 children (weighted) in 617 clusters
  • EDHS 2016 parent survey: two-stage cluster sample across 9 regions and two administrative cities, 21 strata, 645 enumeration areas · human · n = 16,650 households; 15,683 women aged 15-49
Methods & tools
STATA version 12 (svy commands), Ethiopian Demographic and Health Survey 2016, DHS recode-6 manual, WHO infant and young child feeding (IYCF) indicators, variance inflation factor (multicollinearity check)
Reagents & environment
STATA version 12
Reproducibility gaps
No analysis code or scripts are shared. EDHS data were obtained through a written access request to ORC Macro (DHS Program) and kept under a confidentiality agreement.
Limitations (3)
  • Between-community variance remained significant in the combined model (p<0.001). The authors read this as evidence of community-level factors their model did not include. · stated by the authors
  • The EDHS does not record cluster characteristics directly. Community-level variables are therefore proxies: mothers' characteristics averaged within clusters and grouped at national median values. · noted in this summary
  • Effect sizes differ between the abstract and the results (e.g., father's education AOR 2.1 vs 2.4; agrarian region 5.6 vs 5.1). The abstract names urban residence where the results name region, so it is unclear which model is final. · noted in this summary

Read this if you study complementary feeding or child nutrition in Ethiopia, or want a worked example of two-level logistic regression with ICC and variance partitioning on Demographic and Health Survey data.

Citations
80
FWCI
7.97
Field percentile
top 2%

Reproducibility ledger

Nobody has reported trying to reproduce this paper yet. Did you replicate an experiment or run the code? Your report saves the next lab a week.

Sign in to file a reproducibility report.

Related papers

Similarity Map
SelectedSame fieldCross-field
THISmengistie '20Amare '19Bantie '19Gebremariam '20Gebremichae '20liyew '20
74% semantic matchPharmacology & Toxicology
l. a. mengistie, y. worku, W. S. shiferaw+1 more·bioRxiv·Jul 23, 2020