Fig. 4

Evaluation of Model Performance for Aβ-PET Positivity: Influence of eGFR on AUC and AIC. A Area Under the Curve (AUC) values for predicting Aβ-PET positivity using various AD biomarkers across three models: univariate, adjusted for age and sex, and fully adjusted (age + sex + eGFR). The dashed vertical line at 0.50 represents the threshold where the model has no discriminatory ability (essentially, no better than random chance). A higher AUC value indicates better predictive accuracy for Aβ-PET positivity. B ΔAIC values comparing the improvement in model fit by adding eGFR to the age + sex model. The analysis shows that while eGFR adds some predictive value, the improvement in model performance is limited compared to the contribution of age and sex. The dashed vertical line corresponds to the ΔAIC of -2, commonly used in the literature as the minimal difference in AIC to denote statistical significance