Using anthropometric indices to predict cardio-metabolic risk factors in Australian indigenous populations

Abstract 

Aims

To compare the predictive power of anthropometric indices (BMI, waist circumference (WC), waist-to-hip ratio (WHpR) and waist-to-height ratio (WHtR)) for diabetes, hypertension and dyslipidemia in Australian Aboriginal and Torres Strait Islander (TSI) adults.

Methods

Cross-sectional study of 2862 Indigenous Australians aged over 15 living in rural communities in Far North Queensland during 1999–2001. The predictive values of anthropometric indices for cardio-metabolic disorders were compared using receiver operating characteristic (ROC) analysis.

Results

BMI was the poorest predictor while WHpR was the best among the four measures. The optimal WHtR and WHpR cut-off points for the cardio-metabolic risks in both women and men in the two Indigenous populations were 0.5–0.6 and 0.9 respectively. Optimal BMI cut-offs for diabetes, hypertension, and dyslipidemia were much lower in Aborigines than the recommended WHO BMI cut-offs, while those in TSIs were around WHO BMI criteria. The optimal WC cut-points varied by gender and ethnicity.

Conclusions

BMI was not a good discriminator of cardio-metabolic risk factors in Australian Indigenous populations compared with other anthropometric indices. WHpR is more closely associated with the risk of cardio-metabolic in these high-risk populations.

Keywords: Anthropometric indices, Cardio-metabolic risk factors, ROC analysis, Australian Indigenous populations