Fig. 4From: Construction of environmental risk score beyond standard linear models using machine learning methods: application to metal mixtures, oxidative stress and cardiovascular disease in NHANESOdds ratios (95% confidence intervals) of having high GGT (50Â U/L and above) comparing the highest vs. the lowest quintiles of ERS and individual pollutants that compose the ERS in the testing set. All models were adjusted for age, BMI, creatinine, gender, race/ethnicity, smoking status and educationBack to article page