EUVIMEDEUVIMED entdecken
MetaanalyseSystematische ÜbersichtOpen Access · CC BY 4.0

Machine Learning in Predicting Child Malnutrition: A Meta-Analysis of Demographic and Health Surveys Data

Machine Learning zur Vorhersage von kindlicher Mangelernährung: Metaanalyse von Demographic and Health Surveys-Daten

Bhagyajyothi Rao, Muhammad Rashid iD, Md Gulzarull Hasan iD, Girish Thunga

International Journal of Environmental Research and Public Health · 18.03.2025 · 22 (3) · 449

DOI: 10.3390/ijerph22030449 ↗

Auf einen Blick

Die Metaanalyse von 11 Studien zu DHS-Daten zeigte für ML-Modelle eine mittlere bis gute Vorhersage von Mangelernährung bei Kindern unter 5 Jahren. Die gepoolte Genauigkeit variierte je nach Endpunkt.

Fragestellung

Wie gut sagen Machine-Learning-Modelle auf Basis von DHS-Daten Mangelernährung bei Kindern unter 5 Jahren vorher?

Methode

Systematische Suche in PubMed, Embase und Scopus bis Januar 2024; eingeschlossen wurden Studien zu ML-Algorithmen auf DHS-Daten. Qualität mit PROBAST bewertet; Metaanalyse mit Review Manager 5.4.

Ergebnisse

11 von 789 Studien eingeschlossen. Gepoolte Genauigkeit: Stunting 68,92 % (95 %-KI 66,04–71,80; I2 = 100 %), Wasting 84,39 % (95 %-KI 80,90–87,87; I2 = 100 %), Underweight 73,60 % (95 %-KI 70,01–77,20; I2 = 100 %).

Einschränkungen

I2 war bei allen gepoolten Analysen 100 %.

Schlussfolgerung der Autor:innen

Die Metaanalyse zeigte, dass ML-Modelle bei der Vorhersage von Mangelernährung anhand von DHS-Daten bei Kindern unter 5 Jahren mittlere bis gute Leistungswerte aufwiesen.

Kernaussagen

  • 11 von 789 Studien wurden eingeschlossen
  • Gepoolte Genauigkeit: 68,92 % für Stunting
  • Gepoolte Genauigkeit: 84,39 % für Wasting; 73,60 % für Underweight

Deutsche Zusammenfassung: KI-gestützt aus dem Original-Abstract erstellt; alle Zahlen wurden automatisch mit dem Original abgeglichen. Maßgeblich ist der Original-Abstract.

Kurz gefasst

Metaanalyse in der Zeitschrift International Journal of Environmental Research and Public Health, online erschienen am 18.03.2025. Verfasst von Bhagyajyothi Rao und 3 weiteren Autor:innen. Beteiligte Einrichtung: Department of Applied Statistics & Data Science, Prasanna School of Public Health, Manipal Academy of Higher Education, Manipal 576104, India. Hauptthemen laut MEDLINE: Child Nutrition Disorders, Machine Learning. Die Arbeit wurde bisher 32-mal zitiert (Crossref) und nennt 57 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

Auf einen BlickAbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Background

Childhood malnutrition remains a significant global public health concern. The Demographic and Health Surveys (DHS) program provides specific data on child health across numerous countries. This meta-analysis aims to comprehensively assess machine learning (ML) applications in DHS data to predict malnutrition in children.

Methods

A comprehensive search of the peer-reviewed literature in PubMed, Embase, and Scopus databases was conducted in January 2024. Studies employing ML algorithms on DHS data to predict malnutrition in children under 5 years were included. Using PROBAST (Prediction model Risk Of Bias Assessment Tool), the quality of the listed studies was evaluated. To conduct meta-analyses, Review Manager 5.4 was used.

Results

A total of 11 out of 789 studies were included in this review. The studies were published between 2019 and 2023, with the major contribution from Bangladesh (n = 6, 55%). Of these, ten studies reported stunting, three reported wasting, and four reported underweight. A meta-analysis of ten studies reported a pooled accuracy of 68.92% (95% CI: 66.04, 71.80; I2 = 100%) among ML models for predicting stunting in children. Three studies indicated a pooled accuracy of 84.39% (95% CI: 80.90, 87.87; I2 = 100%) in predicting wasting. A meta-analysis of four studies indicated a pooled accuracy of 73.60% (95% CI: 70.01, 77.20; I2 = 100%) for ML models predicting underweight status in children.

Conclusions

This meta-analysis indicated that ML models were observed to have moderate to good performance metrics in predicting malnutrition using DHS data among children under five years.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Child Nutrition DisordersMachine LearningChild, PreschoolHumansInfantHealth Surveys

Stichwörter der Autor:innen

childhood malnutritiondemographic and health surveys (DHS)machine learningmalnutritionmeta-analysispublic healthstunting

Autor:innen und Institutionen

  • Department of Applied Statistics & Data Science, Prasanna School of Public Health, Manipal Academy of Higher Education, Manipal 576104, India.
  • Department of Pharmacy Practice, Manipal College of Pharmaceutical Sciences, Manipal Academy of Higher Education, Manipal 576104, India.
  • Department of Pharmacotherapy, College of Pharmacy, University of Utah, Salt Lake City, UT 84112, USA.
  • Centre for Toxicovigilance and Drug Safety, Manipal College of Pharmaceutical Sciences, Manipal Academy of Higher Education, Manipal 576104, India.

ORCID-Profile: Muhammad Rashid, Md Gulzarull Hasan

Zitieren

APA

Rao, B., Rashid, M., Hasan, M. G., & Thunga, G. (2025). Machine Learning in Predicting Child Malnutrition: A Meta-Analysis of Demographic and Health Surveys Data. International Journal of Environmental Research and Public Health, 22(3), 449. https://doi.org/10.3390/ijerph22030449

Vancouver

Rao B, Rashid M, Hasan MG, Thunga G. Machine Learning in Predicting Child Malnutrition: A Meta-Analysis of Demographic and Health Surveys Data. International Journal of Environmental Research and Public Health. 2025;22(3):449. doi:10.3390/ijerph22030449

BibTeX
@article{Rao2025,
  author = {Rao, Bhagyajyothi and Rashid, Muhammad and Hasan, Md Gulzarull and Thunga, Girish},
  title = {Machine Learning in Predicting Child Malnutrition: A Meta-Analysis of Demographic and Health Surveys Data},
  journal = {International Journal of Environmental Research and Public Health},
  year = {2025},
  volume = {22},
  number = {3},
  pages = {449},
  doi = {10.3390/ijerph22030449}
}
Kostenlos starten →