Multilevel Determinants of Overweight and Obesity Among U.S. Children Aged 10-17: Comparative Evaluation of Statistical and Machine Learning Approaches Using the 2021 National Survey of Children's Health

Joyanta Jyoti Mondal

cs.AIFeb 23, 2026

Multilevel Determinants of Overweight and Obesity Among U.S. Children Aged 10-17: Comparative Evaluation of Statistical and Machine Learning Approaches Using the 2021 National Survey of Children's Health

Authors: Joyanta Jyoti Mondal

Organizations: Department of Computer and Information Sciences, University of Delaware, United States

Abstract

Background: Childhood and adolescent overweight and obesity remain major public health concerns in the United States and are shaped by behavioral, household, and community factors. Their joint predictive structure at the population level remains incompletely characterized. Objectives: The study aims to identify multilevel predictors of overweight and obesity among U.S. adolescents and compare the predictive performance, calibration, and subgroup equity of statistical, machine-learning, and deep-learning models. Data and Methods: We analyze 18,792 children aged 10-17 years from the 2021 National Survey of Children's Health. Overweight/obesity is defined using BMI categories. Predictors included diet, physical activity, sleep, parental stress, socioeconomic conditions, adverse experiences, and neighborhood characteristics. Models include logistic regression, random forest, gradient boosting, XGBoost, LightGBM, multilayer perceptron, and TabNet. Performance is evaluated using AUC, accuracy, precision, recall, F1 score, and Brier score. Results: Discrimination range from 0.66 to 0.79. Logistic regression, gradient boosting, and MLP showed the most stable balance of discrimination and calibration. Boosting and deep learning modestly improve recall and F1 score. No model was uniformly superior. Performance disparities across race and poverty groups persist across algorithms. Conclusion: Increased model complexity yields limited gains over logistic regression. Predictors consistently span behavioral, household, and neighborhood domains. Persistent subgroup disparities indicate the need for improved data quality and equity-focused surveillance rather than greater algorithmic complexity.

Figures & tables

Characteristic	Value
Sample size
Total participants	18,792
Outcome prevalence
Normal weight (%)	67.6
Overweight / Obese (%)	32.4
Demographics

Table 1: Descriptive characteristics of the analytic sample (NSCH 2021, ages 10-17)

Figure 2: Correlation Heatmap of Key Predictors and Outcome.

Panel of Figure 3 — Figure 3: Model stability assessment using cross-validation and learning curves for logistic regression.

Model	AUC	Accuracy	Precision	Recall	F1	Brier
Logistic Regression	0.788	0.769	0.830	0.360	0.502	0.160
Random Forest	0.763	0.765	0.822	0.350	0.491	0.166
Gradient Boosting	0.786	0.765	0.822	0.350	0.491	0.161
XGBoost	0.768	0.761	0.760	0.384	0.510	0.166
LightGBM	0.772	0.768	0.793	0.384	0.518	0.164
Deep Learning (MLP)	0.779	0.766	0.771	0.397	0.524	0.162

Table 2: Predictive performance of all models on the test set. Best values are bolded.

Logistic Regression	LightGBM	TabNet
WgtConcn_21 (2.894)	WgtConcn_21 (0.117)	WgtConcn_21 (0.111)
SC_RACE_R (0.433)	AdultEduc_21 (0.024)	SC_RACE_R (0.082)
AdultEduc_21 (0.382)	SC_AGE_YEARS (0.013)	PhysAct_21 (0.052)
SC_SEX (0.248)	SC_SEX (0.006)	ScreenTime_21 (0.052)
PhysAct_21 (0.181)	SC_RACE_R (0.005)	NbhdAmenities_21 (0.049)
povlev4_21 (0.176)	NbhdSupp_21 (0.003)	NbhdDetract_21 (0.047)

Table 3: Top 10 influential features by model. Values in parentheses indicate importance magnitude: absolute standardized coefficients for Logistic Regression, permutation importance for LightGBM, and aggregated attention importance for TabNet. Bolded features appear consistently across all three models.

Panel of Figure 4 — Figure 4: Top predictors of adolescent overweight and obesity across modeling approaches.

Table 8

Table 9

Tabular ML

Tabular machine learning develops models and tools for learning from structured data arranged in rows and columns.

See all 115 papers

Machine Learning-Based Prediction of Childhood Stunting in Bangladesh: Fairness and Temporal Robustness Assessment
Sep 21, 2026Md Ahshanul Haque, Muhammad Ashad KabirHealthcare Algorithmic Fairness
School of Computing, Mathematics and Engineering, Charles Sturt University, NSW, Australia
Survey-aware Machine Learning: A Guideline for Valid Population Health Inference based on Scoping Review
May 9, 2026YongKyung Oh, Henry W. Zheng, Jeffrey Feng +1
Medical & Imaging Informatics (MII), UCLA
Few-shot Cross-country Generalization of Tabular Machine Learning and Foundation Models for Childhood Anemia Prediction under Distribution Shift
May 26, 2026Yusuf Brima, Marcellin Atemkeng, Lansana Hassim Kallon +4Tabular Foundation Models Few-Shot Learning
Computer Vision, Institute of Cognitive Science, Osnabrück University, Germany · Department of Mathematics, Rhodes University, South Africa · National Institute for Theoretical and computational Sciences (NITheCS), Stellenbosch, 7600, South Africa +6

Sep 21, 2026J/K move · Enter open · S save

Machine Learning-Based Prediction of Childhood Stunting in Bangladesh: Fairness and Temporal Robustness Assessment

Md Ahshanul Haque, Muhammad Ashad Kabir

School of Computing, Mathematics and Engineering, Charles Sturt University, NSW, Australia

Healthcare Algorithmic Fairness Tabular ML

Open paper AskPDF arXiv

Childhood stunting remains a major public health concern in Bangladesh and reflects long-term growth failure influenced by child, maternal, household, socioeconomic, and health-service factors. This study used nationally representative Bangladesh Demographic and Health Survey data from 2007 to 2022 to develop machine learning models for population-level prediction of childhood stunting and to assess temporal robustness and subgroup fairness. Children aged 0-59 months with complete anthropometric and predictor data were included. Data from the 2007, 2011, and 2014 survey rounds were used for model development, while the 2018 and 2022 rounds were retained as temporal test datasets. Twelve feature-selection approaches were assessed, and the KNN permutation importance-selected predictor set was used for final model evaluation. Eleven machine learning models were evaluated: ten conventional algorithms and one pretrained tabular foundation model, TabPFN. Performance was assessed using balanced accuracy, AUROC, F1-score, Brier score, and expected calibration error. Subgroup fairness was examined by child sex, place of residence, and socioeconomic status. The final analytic sample included 18,844 children, of whom 35.05% were stunted. In the development hold-out test dataset, TabPFN showed the highest observed balanced accuracy overall at 67.58%, while AdaBoost showed the highest observed balanced accuracy among conventional models at 67.51%. In temporal testing, the highest observed balanced accuracy was found for Gradient Boosting in BDHS 2018 and XGBoost in BDHS 2022. Model performance varied across survey rounds and subgroups, highlighting the importance of temporal validation, subgroup fairness assessment, and transparent interpretation in public health prediction modeling.

Abstract

Tabular ML

Explore similar work

Machine Learning-Based Prediction of Childhood Stunting in Bangladesh: Fairness and Temporal Robustness Assessment

Survey-aware Machine Learning: A Guideline for Valid Population Health Inference based on Scoping Review

Few-shot Cross-country Generalization of Tabular Machine Learning and Foundation Models for Childhood Anemia Prediction under Distribution Shift