Background: Optimal cutoffs for obesity indices are rarely studied in Asians. We evaluated these cutoffs for diabetes, hypertension, dyslipidemia and any risk factor for the Taiwanese general population. Methods: Body mass index (BMI), waist circumference (WC), waist-hip ratio (WHR), waist-height ratio (WHeiR) and other data for 4683 (2280 men and 2403 women) participants of the population-based Taiwanese Survey on Hypertension, Hyperglycemia and Hyperlipidemia were used. Areas under curves (AUCs) were analyzed and optimal cutoffs were estimated by maximizing the sums of sensitivity and specificity. Potential confounders included age, smoking, alcohol, betel nut chewing and exercise. Results: Optimal cutoffs for men and women, respectively, were 23.7-26.3 and 22.1-23.2kg/m2 for BMI; 85.0-87.0 and 74.0-83.0cm for WC; 0.87-0.90 and 0.78-0.83 for WHR; and 0.48-0.52 and 0.48-0.52 for WHeiR. AUCs were not significantly different among the indices for diabetes in men and for hypertension in women. In men, WHR was significantly inferior to the other indices for predicting hypertension, dyslipidemia and any risk factor. In women, BMI was significantly inferior to the others for diabetes. For dyslipidemia and any risk factor in women, WHeiR showed the largest AUCs and significant differences were seen in the following pairs: WHeiR vs. BMI and WHeiR vs. WHR for dyslipidemia and WC vs. WHR and WHeiR vs. WHR for any risk factor. Conclusions: WC and WHeiR have similar efficacy and are superior to BMI and WHR. However, WHeiR has the extra benefit of a unisex cutoff within a narrow range.

Original languageEnglish
Pages (from-to)585-589
Number of pages5
Issue number2
Publication statusPublished - Jun 2010


  • Anthropometric factor
  • Diabetes
  • Dyslipidemia
  • Hypertension
  • Obesity

ASJC Scopus subject areas

  • Cardiology and Cardiovascular Medicine


Dive into the research topics of 'Optimal anthropometric factor cutoffs for hyperglycemia, hypertension and dyslipidemia for the Taiwanese population'. Together they form a unique fingerprint.

Cite this