Objective: To evaluate the validity of currently recommended obesity cutoffs of body mass index (BMI, in kg/m(2)) and waist circumference (WC, in cm) for Asians by the WHO/IASO/IOTF and for Chinese by the Working Group on Obesity in China (WGOC) using the percentage body fat (%BF)-obesity criteria.

Design: A cross-sectional study.

Subjects: A total of 1122 community-based Hong Kong Chinese women aged between 41 and 63 years.

Measurements: Total %BF and percent truncal fat (%TF) were measured using dual-energy X-ray absorptiometry. Anthropometric indices were measured using standard methods.

Results: Regression analyses showed that the BMI cutoffs of 23, 24, 25, and 28 kg/m(2) corresponded to the %BF of 34.8, 35.9, 36.9 and 39.5%, and the 80 cm WC corresponded to 34% TF, respectively. Compared with the %BF obesity cutoff (>/=40%), the WHO/IASO/IOTF BMI-obesity criterion (>/=25) shows a good sensitivity (75%) and specificity (71%); and the WGOC criterion (BMI>/=28) had a low sensitivity (41%) but an excellent specificity (93%), respectively. Corresponding to the BMI cutoffs of 23, 24, 25 and 28 kg/m(2), the %BF cutoffs associated with peak kappa statistic were 33, 34, 35 and 40%, and the relevant %TF linked with 80 cm WC was 33%, respectively.

Conclusion: BMI and WC have a good accuracy in the prediction of obesity. Our findings suggest that the WGOC BMI cutoffs are appropriate, but 80 cm of WC is a very rigorous cutoff for this population when using the criteria of 34 and 40% of body fat or truncal fat for overweight and obesity.