Objective  To assess relationships between obesity indexes and hypertension and to identify strong obesity indexes for predicting hypertension risk in older adults.

  Methods  We extracted physical measurement data on 1 608 community residents aged 60 – 69 years from the dataset of two waves of the National Physical Fitness Surveillance conducted in 2010 and 2014 in Liaoning province. We calculated conventional obesity indexes including body mass index (BMI), ponderal index (PI), waist circumference (WC), waist-to-hip ratio (WHR), and conicity index (CI) for all the partipants and some other novel obesity indexes such as waist-to-height ratio (WHtR), a body shape index (ABSI), body roundness index (BRI), abdominal volume index (AVI), a new hip index (HI), and body adiposity index (BAI) were also calculated simultaneously. Receiver operating characteristic (ROC) curve was applied to compare the power of different obesity index in predicting hypertension and logistic regression analysis was used to evaluate the strength of association between obesity indexes and hypertension risk.

  Results  The prevalence of hypertension was 38.6% (620/1 608) for all the participants and was 41.5% (323/778) and 35.8% (297/830) for the male and female participants. Significant prediction of hypertension using application of area under ROC curve in predicting hypertension was of statistically significance for the indexes of BMI, WC, WHtR, AVI, PI, BRI and BAI in both male and female participants, with the optimal cut-off values of 25.4 and 25.6 kg/m² (for male and female) for BMI, 83.5 and 86.6 cm for WC, 0.50 and 0.53 for WHtR, 14.3 and 15.2 cm² for AVI, 14.6 and 16.1 kg/m³ for PI, 3.4 and 4.1 for BRI, and 25.4 and 32.6 for BAI, respectively. The results of multivariate logistic regression analysis demonstrated that BMI and WHtR were good predictors of hypertension risk; among the males, the risk of hypertension for the participants with the BMI ≥ 25.4 kg/m² was 1.53 times higher than those with the BMI < 25.4 kg/m² and the risk of hypertension for the participants with the WHtR ≥ 0.50 was 1.55 times higher than those with the WHtR < 0.50; among the females, the risk of hypertension for the participants with the BMI ≥ 25.6 kg/m² was 1.62 times higher than those with the BMI < 25.6 kg/m² and the risk of hypertension for the participants with the WHtR ≥ 0.53 was 1.81 times higher than those with the WHtR < 0.53; the risk of hypertension for the male participants in men with the BMI ≥ 25.4 kg/m² and the WHtR ≥ 0.50 was 2.33 times (95% confidence interval 95% CI: 1.63 – 3.34) higher than the males with the BMI < 25.4 kg/m² and the WHtR < 0.50, while, for the female participants, the risk of hypertension in the females with the BMI ≥ 25.6 kg/m² and the WHtR ≥ 0.53 was 2.88 times (95% CI: 2.03 – 4.09) higher than those with the BMI < 25.6 kg/m² and the WHtR < 0.53.

  Conclusion  Body mass index and waist-to-height ratio are good obesity indexes for predicting hypertension risk in elderly community residents and the prediction power could be significantly improved when the two indexes being used jointly.