Background: Hypertension is related to increased body fat, which can be evaluated by anthropometric indicators.
Objective: To determine the predictive power of anthropometric indicators and establish their cutoff points as discriminators of high blood pressure.
Methods: Cross-sectional study with a sample of 660 adolescents aged 14 to 19 including 51.9% girls. We considered the following anthropometric indicators: body mass index (BMI), waist circumference, waist-to-height ratio and conicity index. High blood pressure was characterized by values above the 90th percentile for systolic and/or diastolic blood pressure. To identify predictors of high blood pressure, we adopted the analysis of receiver operating characteristic curves (ROC) with a confidence interval of 95%. Subsequently, we identified the cutoff points with their relevant sensitivities and specificities.
Results: The areas under the ROC curves with confidence intervals were: boys--waist circumference = 0.80 (0.72 to 0.89); BMI = 0.79 (0.68 to 0.89), waist-to-height ratio = 0.77 (0.66 to 0.88); conicity index = 0.69 (0.56 to 0.81) and for girls--waist circumference = 0.96 (0.92 to 1.00); BMI 0.95 (0.87 to 1.00), waist-to-height ratio = 0.93 (0.85 to 1.00); conicity index = 0.74 (0.50 to 0.98). The different cutoff points of anthropometric indicators with better predictive power and their relevant sensitivities and specificities were identified.
Conclusion: Although the waist-to-height ratio and BMI have shown good areas under the ROC curve, we suggest the use of waist circumference to predict high blood pressure.