Skip to main content
Fig. 3 | Lipids in Health and Disease

Fig. 3

From: Development and application of a novel model to predict the risk of non-alcoholic fatty liver disease among lean pre-diabetics with normal blood lipid levels

Fig. 3

Selection of variables using the least absolute shrinkage and selection operator (LASSO) binary logistic regression model. A coefficient profile plot was constructed against the log (lambda) sequence. A Seventeen variables with nonzero coefficients were selected by deriving the optimal lambda value. B Following verification of the optimal parameter (lambda) in the LASSO model, partial likelihood deviance (binomial deviance) curve versus log (lambda) was plotted, and dotted vertical lines for those variables were drawn, based on 1 standard error criteria, to obtain the 7 variables (body mass index [BMI], total cholesterol [TC], alanine aminotransferase to aspartate aminotransferase ratio [AAR], triglyceride to high density lipoprotein cholesterol ratio [THR], fasting blood glucose [FPG], γ-glutamyl-transferase [GGT], and uric acid [UA]). Construction of the predictive nomogram. C The predictive nomogram is based on the risk factors of GGT, AAR, FPG, BMI, THR, TC, and BMI. D Example of a nomogram in use, where the patient measurements for each of the 6 parameters corresponds to a specific point value, and the total points corresponds to a percentage likelihood of developing NAFLD

Back to article page

Lipids in Health and Disease

ISSN: 1476-511X