Statistical analysis
All analyses were performed with the IBM Statistical Package for the Social Science version 25.0 (SPSS Inc., Chicago, IL, USA) for MacOS. Descriptive data were given as percentages and as median (IQR 25-75). The Spearman’s correlation test was used to confirm the correlations between CG versus (vs.) serum immunoglobulins and gamma globulin vs. serum immunoglobulins. The power of correlation was defined as very weak if r <0.2, weak if r=0.2–0.4, moderate if r=0.4–0.6, strong if r=0.6–0.8 and very strong if r>0.8.
Receiver operating characteristic (ROC) curve analyses were performed to verify if CG and gamma globulin were independent discriminative values to detect subjects with IgG serum levels lower than 600, 500, 400 and 300 mg/dL. The accuracy of the obtained discriminant value was interpreted based on the area under curve (AUC). The Youden’s index was calculated to confirm the discriminant score, defined as the highest value observed for the following operation: sensitivity + specificity – 110. Since CG was designed as a screening test, the cut-off value which has a sensitivity of >80% was accepted as the most appropriate among the values with the highest Youden’s index10,11.
The chi-square test was applied to compare the frequency of different clinical characteristics between the CG value of <2.0 and ≥2.0. Significant variables were further evaluated using the logistic regression analysis. The results were assessed at a significant level of p<0.05 and a 95% confidence interval (CI).