Table 3_Uncovering non-linear dietary predictors of cardiovascular disease risk in older a

Data from three NHANES cycles (2009–2014) and the MyPyramid Equivalents Database (MPED) were integrated to investigate non-linear dietary predictors of cardiovascular disease risk in older adults with periodontitis. The dataset was used to construct six machine learning models, with an external validation cohort of 183 older adults. The dataset was authored by Jiefan Liu and last updated on March 18, 2026.

Use Cases

• Training interpretable machine learning models (e.g., XGBoost) to predict CVD risk based on dietary trace elements and food intake.
• Applying SHAP and LIME algorithms to identify key protective dietary factors, such as theobromine, lycopene, and meat consumption, in a specific inflammatory population.
• Validating non-linear dietary associations through external cohort analysis using multivariable logistic regression.
• Investigating how periodontitis-induced inflammation may alter traditional dietary risk factors for cardiovascular disease.

Strengths

• Includes an external validation cohort of 183 older adults with periodontitis from an independent hospital-based cohort.
• Model validation reported high predictive performance (AUC-ROC: 0.854 for NHANES, 0.889 for MPED).
• Feature selection was performed using correlation coefficients and the BORUTA algorithm.
• Results include specific odds ratios from logistic regression (e.g., red meat OR=0.46, sweets OR=0.11).

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• The dataset is very small (13.4 KB), indicating limited raw data or a summary document.

Provenance

Source

NHANES cycles (2009–2014) and MyPyramid Equivalents Database (MPED), with external validation from an independent hospital-based cohort.

Collection Method

Cross-sectional analysis integrating survey and database data, with machine learning model construction and validation.

Time Range

2009–2014 (NHANES cycles).

Freshness

Last updated 2026-03-18 06:42:30; freshness should be verified.

Geography

Likely United States (based on NHANES source), but specific geography is not stated.