Is central obesity a better discriminator of the risk of hypertension than body mass index in ethnically diverse populations?

J Hypertens. 2008 Feb;26(2):169-77. doi: 10.1097/HJH.0b013e3282f16ad3.

Abstract

Objective: To compare the performance of body mass index (BMI) against waist circumference, waist: hip ratio (WHR) and waist: height ratio in the discrimination of hypertension in ethnically diverse populations.

Methods: Meta-analysis of 19 cross-sectional studies.

Main outcome measures: Discrimination of hypertension (SBP/DBP > or = 140/90 mmHg) was adjudicated from Receiver Operating Characteristic curves; optimum thresholds were defined as those that maximized sensitivity plus specificity.

Results: Irrespective of which measure of overweight was used, the strength of the association with blood pressure was consistently greater among Asians compared with Caucasians or Pacific Islanders; however, in all regions, and for all anthropometric measures, the increment in blood pressure, and the additional risk of hypertension, were broadly similar for the same relative increment in each of the four measures. Optimum thresholds varied by region; WHR was the most consistent between the regions, with thresholds of 0.92-0.94 for men and 0.80-0.88 for women. No anthropometric variable was systematically better than others at the discrimination of hypertension.

Conclusions: Blood pressure is similarly associated with each of the four measures of overweight chosen, but the associations were stronger among Asians. WHR has advantages in terms of consistency of thresholds for hypertension across ethnic groups in the Asia-Pacific.

Publication types

  • Meta-Analysis
  • Research Support, Non-U.S. Gov't

MeSH terms

  • Adult
  • Area Under Curve
  • Asian People
  • Body Mass Index
  • Female
  • Humans
  • Hypertension / epidemiology
  • Hypertension / ethnology*
  • Male
  • Middle Aged
  • Native Hawaiian or Other Pacific Islander
  • Obesity / complications*
  • Odds Ratio
  • Reference Values
  • Risk
  • Waist-Hip Ratio / adverse effects*
  • White People