Comparison of the ability to identify arterial stiffness between two new anthropometric indices and classical obesity indices in Chinese adults

Atherosclerosis. 2017 Aug:263:263-271. doi: 10.1016/j.atherosclerosis.2017.06.031. Epub 2017 Jun 16.

Abstract

Background and aims: The association between anthropometric indices of body composition and arterial stiffness is inconclusive. The objective of this study was to examine the predictive ability of two new obesity indices: a body shape index (ABSI), and a body roundness index (BRI), for the identification of arterial stiffness among Chinese adults, as well as to compare the relative strength of association between the anthropometric indices and arterial stiffness.

Methods: A total of 10,197 subjects were recruited in this cross-sectional study. We tested the association between anthropometric indices (body mass index [BMI], waist circumference [WC], waist-height-ratio [WHtR], ABSI and BRI) and brachial-ankle pulse wave velocity (PWV). Receiver operating characteristic curve and area under curve (AUC) were employed to evaluate the predictive value of the anthropometric indices for identification of arterial stiffness.

Results: After adjusting for confounding variables, BRI (linear regression: 0.112; AUC: 0.726; OR: 1.228 for female and linear regression: 0.047; AUC: 0.631; OR: 1.173 for male) exhibited a more powerful predictive ability of arterial stiffness than ABSI (linear regression: 0.110; AUC: 0.674; OR: 1.315 for female and linear regression: 0.058; AUC: 0.610; OR: 1.150 for male) and WC (linear regression: 0.078; AUC: 0.699; logistic regression: negative for female and linear regression: negative; AUC: 0.593; logistic regression: negative for male) while having a similar predictive value to that of WHtR (linear regression: 0.113; AUC: 0.726; OR: 1.228 for female and linear regression: 0.047; AUC: 0.631; OR: 1.185 for male) among both sexes. BMI (linear and logistic regression: negative; AUC: 0.660 for female and 0.568 for male) had the lowest predictive power in both sex categories. The optimal cut-off of WHtR for detecting arterial stiffness was 0.49 in females and 0.53 in males, that of BRI was 3.19 in females and 3.89 in males.

Conclusions: WHtR, ABSI and BRI were significantly associated with arterial stiffness. BRI and WHtR, rather than ABSI, showed superior predictive abilities for arterial stiffness in both sexes.

Keywords: Anthropometric index; Arterial stiffness; Brachial-ankle pulse wave velocity; Cardiovascular disease.

Publication types

  • Comparative Study

MeSH terms

  • Adiposity*
  • Adolescent
  • Adult
  • Aged
  • Aged, 80 and over
  • Ankle Brachial Index
  • Anthropometry / methods*
  • Area Under Curve
  • Asian People
  • Body Mass Index
  • Cardiovascular Diseases / diagnosis*
  • Cardiovascular Diseases / epidemiology
  • Cardiovascular Diseases / physiopathology
  • Chi-Square Distribution
  • China
  • Cross-Sectional Studies
  • Female
  • Humans
  • Linear Models
  • Logistic Models
  • Male
  • Middle Aged
  • Obesity / diagnosis*
  • Obesity / epidemiology
  • Obesity / physiopathology
  • Odds Ratio
  • Predictive Value of Tests
  • Pulse Wave Analysis
  • ROC Curve
  • Reproducibility of Results
  • Risk Factors
  • Sex Factors
  • Vascular Stiffness*
  • Waist Circumference
  • Waist-Height Ratio
  • Young Adult