Abstract
Background: The suggestion that body mass index (BMI) cutoffs to define obesity should differ in persons of Polynesian descent compared with Europeans is based principally on the observation that persons of Polynesian descent have a relatively higher proportion of lean body mass for a given BMI.
Objectives: The objectives were to determine whether the relation between BMI, waist circumference, and metabolic comorbidity differs in the 2 major ethnic groups in New Zealand and to ascertain whether ethnicity-specific BMI and waist circumference cutoffs for obesity are justified for Maori (indigenous New Zealanders).
Design: Subjects included a convenience sample of 1539 men and women aged 17 82 y (47% Maori, 53% white) with measures of BMI, waist circumference, blood pressure, fasting insulin, glucose, and lipids. The sensitivity and specificity of BMI (in kg/m(2); 30 and 32), waist circumference (80 and 88 cm in women, 94 and 102 cm in men), and waist-to-height ratio (WHtR; >= 0.6) in relation to insulin sensitivity, insulin resistance, and the metabolic syndrome were determined. Receiver operating characteristic curves and areas under the curve (AUCs) were also calculated.
Results: No ethnic or sex differences between AUCs were observed for BMI, waist circumference, or WHtR, which showed that these anthropometric measures perform similarly in Maori and European men and women and correctly discriminate between those with and without insulin resistance or the metabolic syndrome 79-87% of the time. Any increase in specificity from a higher BMI cutoff of 32 in Maori was offset by appreciable reductions in sensitivity.
Conclusion: These findings argue against having different BMI or waist circumference cutoffs for people of Polynesian descent. Am J Clin Nutr 2010; 92: 390-7.