Accuracy of body mass index in detecting an elevated alanine aminotransferase level in adolescents.

AIM We evaluated the accuracy of body mass index (BMI) in detecting an elevated alanine aminotransferase (ALT) level in adolescents, taking into account the effects of gender, age, ethanol intake, hepatitis B virus (HBV) and hepatitis C virus (HCV) infections, and drug consumption. SUBJECTS A representative sample of 454 adolescents (11-17 years) from two cities in northern Italy was studied (the Dionysos Study). METHODS z-BMI was calculated as the z-score of BMI using national growth charts. Logistic regression was used to quantify the contribution of the variables of interest to an elevated ALT (> 30 UL(-1)). Odds ratios (OR) and 95% confidence intervals (95% CI) were calculated, and areas under receiver-operator characteristic curves (AUC) were used to evaluate accuracy. RESULTS An elevated ALT was detected in 21 adolescents (4.6%). Among the studied variables, only male gender (OR=6.7, 95% CI 2.0-23.2) and z-BMI (OR=2.1, 95% CI 1.4-3.2) were significant predictors of elevated ALT. The accuracy of the prediction was 0.69 (95% CI 0.59-0.79) for gender and 0.71 (95% CI 0.59-0.81) for z-BMI. By combining gender and z-BMI, the accuracy rose to 0.80 (95% CI 0.71-0.89). CONCLUSION BMI is a good predictor of elevated ALT in Italian adolescents and gender adds to the accuracy of the prediction.