Which anthropometric indices can best estimate obesity amongst Ellisras population from childhood to young adulthood? Ellisras Longitudinal Study

Abstract

Abdominal and general obesity are considered to be predisposing factors for several cardiovascular diseases such as hypertension. This study examined the most appropriate anthropometric tool for estimating obesity among Ellisras population from childhood to young adulthood. Using a crosssectional design, the study measured the anthropometric characteristics of 10,811 (aged 3 -9 years) and 14,311 children (aged 10-16 years) as well as 770 young adults (aged 18- 29 years). Accuracy of the various obesity measures (general and abdominal obesity) was evaluated using Receiver Operating Characteristic (ROC) analysis. Females showed higher (51.3%) significant (p<0.05) prevalence of abdominal obesity compared to males (4.3%) at an older age. ROC analysis revealed waist circumference (WC) (Area under the ROC curve: AUC= 0.974) and waist to height ratio (WHtR) (AUC=1.000) as the most appropriate tools to classify abdominally obese subjects at an older age. All the central obesity indices significantly (p<0.001) predicted the risk of overall obesity, the best being WC and WHtR followed by waist-hip ratio (WHR) both when unadjusted (B=0.431-63.417) and adjusted (0.350-2.818) for age and gender. Appropriate anthropometric tools for measuring general and abdominal obesity among the population in this setting for children are WC and body mass index (BMI) and WHtR and WC for young adults.

Keywords: Obesity, prevalence, children, young adults, anthropometry.