In this study we compared the diagnostic utility of several anthropometric and body composition measures in predicting elevated fasting blood sugar (FBS). Participant had overnight FBS analyzed with a hand‐held automated glucometer (Basic One‐touch Ultra). Waist circumference (WC), height (Ht) and weight(Wt)were measured using standard procedure. These were used to calculate BMI and waist circumference‐to‐height ratio (WHr). Percentage body fat (%BF) was measured by foot‐to‐foot bioelectric impedance (Tanita UM026).To determine diagnostic utility of the various measures, sensitivities, specificities and area under the operator receiver curve (AUC) were computed using recommended cut‐off values. 157 persons participated in the study (females = 90; males = 67). FBS was significantly correlated with Wt, WC, fat mass, and BMI in both groups. %BF was significantly associated with FBS in females only. Sensitivities for the various anthropometric and body composition cut‐off values range from 53%‐85% in males and 61%‐100% among females. Specificities range from 45%‐88% in males and 42%‐79% among females. AUC analyses suggest that BMI>25, WC, WHr were be superior indices for predicting elevated FBS among females only. Our results suggest that among participants several indices were useful in predicting FBS, however, this ability to predict was gender specific.

No abstract