Key result
Machine learning models show similar discrimination for identifying hypertension, led by logistic regression with ~0.79 ROC-AUC.
Why the study?
Hypertension is a major noncommunicable disease and public health concern, prompting this study to compare machine learning models, hyperparameter optimization, and predictor contributions for classifying reported hypertension status.
Do tuned Random Forest and XGBoost machine learning models improve the classification of self-reported physician-diagnosed hypertension status compared to baseline Logistic Regression in adult NHANES 2021-2023 respondents?
Cross-Sectional (n=5,008)
Yes
Do tuned Random Forest and XGBoost machine learning models improve the classification of self-reported physician-diagnosed hypertension status compared to baseline Logistic Regression in adult NHANES 2021-2023 respondents?
In classifying self-reported hypertension using basic demographic and anthropometric NHANES data, standard logistic regression provided comparable overall discrimination (ROC-AUC) to tuned tree-based machine learning models, while SHAP analysis confirmed age, waist circumference, and BMI as consistently dominant predictors across ensemble architectures.
Tuned random forest may aid hypertension classification in survey data; leaves open superiority to logistic regression in clinical settings.
Hypertension is a major noncommunicable disease and an important public health concern. Its health burden is relevant to public-health risk management and broader disaster-risk contexts, supporting Sustainable Development Goal 3 (SDG 3), particularly Target 3.4 on reducing premature mortality from noncommunicable diseases. This study compares Logistic Regression (LR), Random Forest (RF), and Extreme Gradient Boosting (XGBoost) for classifying reported hypertension status, evaluates hyperparameter optimization, and examines predictor contributions using SHapley Additive exPlanations (SHAP). Using 5,008 NHANES 20212023 respondents and eight predictors, hypertension status was defined from BPQ020 as self-reported physician diagnosis. The data were divided using an 80:20 stratified split into training and independent test sets, with five-fold cross-validation applied to the training set for hyperparameter optimization. At a fixed threshold of 0.50, LR achieved the highest baseline ROC-AUC (0.7921), while XGBoost achieved the highest baseline Recall (0.5968) and F1-Score (0.6233). Tuning improved RF Recall from 0.5332 to 0.6048 and F1-Score from 0.5793 to 0.6316. SHAP identified Age, Waist circumference, and BMI as the top three predictors in both tuned models, with a Spearman rank correlation of 0.9286 and 100% top-three agreement. Bootstrap analysis yielded a median correlation of 0.9286 (95% CI: 0.90481.0000), with 100% top-three agreement across 2,000 replications. At the fixed threshold, tuned RF achieved the highest Recall and F1-Score, whereas LR yielded the highest ROC-AUC.
No takes yet. Share an insight, caveat, or question.
Apriani et al. (2026) conducted a cross-sectional in Hypertension (n=5,008). Tuned Random Forest and Tuned XGBoost classification models vs. Logistic Regression baseline model was evaluated on Classification of self-reported physician-diagnosed hypertension status (ROC-AUC) (95% CI 0.7543-0.8089, p=0.6513).
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: