The present study was performed to test the predictive value of different cut-off points of anthropometric parameters for the presence of type 2 diabetes (T2DM) or glucose tolerance abnormalities in north-east Chinese adults.

Multistage random cluster sampling method in a cross-sectional study.

Height, body weight, maximal body weight in the past, waist and hip circumferences, blood pressure, 2 h post-load glucose and other lifestyle factors were measured.

We used data from 1058 adults aged 20 years or over, selected in the city of Mudanjiang, in 2005.

BMI, maximal BMI in the past (MAXBMI), waist:hip ratio (WHR), waist:height ratio (WHtR) and waist circumference (WC) were significantly correlated with each other. Partial correlation coefficients between WHtR and WC, and between MAXBMI and BMI, were higher than those between the other indices. The association of anthropometric indices with T2DM or glucose tolerance abnormalities was significantly highest for the collaboration cut-off points of MAXBMI (≥23·0 kg/m2 for T2DM, ≥22·0 kg/m2 for glucose tolerance abnormalities) with WHtR (≥0·52). Areas under the receiver-operating characteristic curves also showed that WHtR was a better anthropometric index that discriminated between the presence and absence of T2DM and an excellent indicator with high Youden’s index.

MAXBMI combined with WHtR was a better anthropometric index associated with T2DM or glucose tolerance abnormalities. The combined use of these two measures is a good choice for T2DM prevention and screening.