High-density lipoprotein-related inflammatory ratios and coronary heart disease: A cross-sectional machine learning analysis of NHANES 2009 to 2020.

Self-reported CHD prevalence in the study sample was 5.7%.

• 840 of 14,745 US adults self-reported a CHD diagnosis.
• The sample included adults aged ≥20 years with a mean age of 51.8 ± 17.6 years.
• Data were drawn from NHANES 2009 to 2020.

All HDL-related inflammatory ratios were significantly elevated in CHD patients compared to non-CHD participants.

• MHR was 0.54 ± 0.35 in CHD patients vs 0.42 ± 0.23 in non-CHD participants (P < .001).
• Lymphocyte-to-HDL cholesterol ratio was 2.05 ± 3.12 vs 1.55 ± 1.02 (P < .001).
• NHR was 4.06 ± 2.89 vs 3.11 ± 1.77 (P < .001).
• The four ratios examined were MHR, lymphocyte-to-HDL cholesterol ratio, NHR, and platelet-to-HDL cholesterol ratio.

The eXtreme gradient boosting (XGBoost) model demonstrated optimal performance among the machine learning models tested.

• XGBoost achieved an area under the receiver operating characteristic curve (AUROC) of 0.8892.
• Accuracy was 96.55% and precision was 86.00%.
• Three machine learning models were compared: eXtreme gradient boosting, random forest, and logistic regression.

SHAP (SHapley Additive exPlanations) analysis identified age as the most important predictor of CHD, with MHR and NHR ranking among the top 5 features.

• SHAP analysis was used to provide model interpretability for the XGBoost model.
• MHR and NHR were among the top 5 features in the SHAP-ranked feature importance.
• Age was ranked as the single most important predictor variable.

The study identified cross-sectional associations between HDL-related inflammatory ratios and CHD prevalence rather than predictive relationships for incident events.

• The study design was cross-sectional using NHANES 2009 to 2020 data.
• The outcome variable was self-reported CHD diagnosis.
• The authors explicitly noted these findings 'reveal significant cross-sectional associations between HDL-related inflammatory ratios and CHD prevalence, rather than predictive relationships for incident events.'
• Prospective validation was identified as warranted to establish utility for predicting incident CHD events.

HDL-related inflammatory ratios are composite biomarkers derived from routine blood tests that integrate lipid metabolism and inflammatory pathways.

• The four ratios studied were monocyte-to-HDL cholesterol ratio (MHR), lymphocyte-to-HDL cholesterol ratio, neutrophil-to-HDL cholesterol ratio (NHR), and platelet-to-HDL cholesterol ratio.
• These biomarkers are described as 'readily available biomarkers from routine blood tests' that 'provide substantial value for cardiovascular risk stratification.'
• The study population comprised 14,745 US adults aged ≥20 years.