Identification of BMI-related high-risk feature combinations for diabetes among young adults with normal baseline fasting plasma glucose using interpretable machine learning: a health check-up cohort study.

It should be noted that although total cholesterol ranked highly in the SHAP importance analysis, its association did not reach statistical significance in the multivariable Cox model; therefore, it is more appropriate to describe it as a variable with high contribution in the exploratory machine learning model rather than as an independent risk factor.

also in 111,027 other papers

The BMI ≥28 × TG ≥1.7 mmol/L interaction term did not reach statistical significance but also showed a suggestive trend (HR = 0.644, 95% CI: 0.381–1.089, P for interaction = 0.105).

also in 203 other papers