Objectives: There is no agreed-upon definition for severe obesity (Sev-OB) in children. We compared estimates of Sev-OB as defined by different cut-points of body mass index (BMI) from the Centers for Disease Control and Prevention (CDC) or the World Health Organization (WHO) curves and the ability of each set of cut-points to screen for the presence of cardiometabolic risk factors.
Research design and methods: Cross-sectional, multicenter study involving 3,340 overweight/obese young subjects. Sev-OB was defined as BMI ≥ 99(th) percentile or ≥ 1.2 times the 95(th) percentile of the CDC or the WHO curves. High blood pressure, hypertriglyceridemia, low High Density Lipoprotein -cholesterol and impaired fasting glucose were considered as cardiometabolic risk factors.
Results: The estimated prevalence of Sev-OB varied widely between the two reference systems. Either using the cut-point ≥ 99(th) percentile or ≥ 1.2 times the 95(th) percentile, less children were defined as Sev-OB by CDC than WHO (46.8 vs. 89.5%, and 63.3 vs. 80.4%, respectively p<0.001). The CDC 99(th) percentile had lower sensitivity (58.5 vs 94.2), higher specificity (57.6 vs 12.3) and higher positive predictive value (34.4 vs 28.9) than WHO in identifying obese children with ≥ 2 cardiometabolic risk factors. These differences were mitigated using the 1.2 times the 95(th) percentile (sensitivity 73.9 vs. 88.1; specificity 40.7 vs. 22.5; positive predictive value 32.1 vs. 30.1). Substantial agreement between growth curves was found using the 1.2 times the 95(th) percentile, in particular in children ≤ 10 years.
Conclusions: Estimates of Sev-OB and cardiometabolic risk as defined by different cut-points of BMI are influenced from the reference systems used. The 1.2 times the 95(th) percentile of BMI of either CDC or WHO standard has a discriminatory advantage over the 99(th) percentile for identifying severely obese children at increased cardiometabolic risk, particularly under 10 years of age.