Body roundness index (BRI) and waist-to-height ratio (WHtR) exhibited the highest discriminatory ability for hypertension among 10 obesity-related indicators in Chinese adults, with both showing nonlinear dose-response relationships and optimal cut-off values of 3.66 and 0.52, respectively.
Key Findings
Results
The weighted prevalence of hypertension among Chinese adults in this national sample was 24.79%.
Sample comprised 117,609 adults aged 20-79 years from the 2020 National Physical Fitness Surveillance Database
All participants had complete data on 10 obesity-related indicators (ORIs) and relevant covariates
The study used a cross-sectional design with nationally representative data
Results
BRI and WHtR demonstrated the highest discriminatory ability for hypertension among the 10 obesity-related indicators evaluated.
AUC for overall population: 0.686 for both BRI and WHtR
AUC in males: 0.647; AUC in females: 0.721
10 ORIs were compared using ROC curves and Youden's Index
Performance was consistent across sex-stratified subgroups
Results
The optimal cut-off values for identifying hypertension risk were BRI ≥ 3.66 and WHtR ≥ 0.52 in the overall population.
Cut-off values were determined using Youden's Index (sensitivity + specificity - 1)
Individuals at or above BRI threshold had a 2.00-fold increased possibility of hypertension compared to those below
Individuals at or above WHtR threshold had a 1.97-fold increased possibility of hypertension compared to those below
Sensitivity analyses substantiated the robustness of these findings
Results
Each unit increase in BRI z-score was associated with a 51% increased odds of hypertension in the overall population.
OR = 1.51 (95% CI: 1.48–1.53) per unit increase in BRI z-score
Estimates derived from multivariable logistic regression models
Results were adjusted for relevant covariates
Results
Each unit increase in WHtR z-score was associated with a 53% increased odds of hypertension in the overall population.
OR = 1.53 (95% CI: 1.50–1.55) per unit increase in WHtR z-score
Estimates derived from multivariable logistic regression models
WHtR z-score showed marginally stronger association than BRI z-score in the overall population
Results
BRI and WHtR both showed nonlinear dose-response relationships with hypertension.
Nonlinear relationships assessed using restricted cubic spline (RCS) models
P for nonlinearity < 0.05 for both BRI and WHtR
The nonlinear pattern suggests the relationship between abdominal obesity and hypertension risk is not constant across the range of values
Conclusions
BRI and WHtR, as abdominal obesity indices incorporating both waist circumference and height, showed highly consistent performance with each other across all analyses.
Both indices incorporate waist circumference and height simultaneously
Performance was consistent in overall population and sex-stratified subgroups
Sensitivity analyses substantiated the robustness of both discriminatory and associative findings
BRI values below 3.66 and WHtR values below 0.52 were identified as potentially conferring substantial reduction in hypertension possibility
What This Means
This research used data from over 117,000 Chinese adults to compare 10 different ways of measuring obesity and determine which best identifies people at risk for high blood pressure (hypertension). The study found that about 1 in 4 Chinese adults (24.79%) had hypertension. Among all the obesity measures tested, two body shape indices — the Body Roundness Index (BRI) and the waist-to-height ratio (WHtR) — were the best at distinguishing people with hypertension from those without, and they performed similarly to each other. Both measures use waist circumference and height together, capturing abdominal (belly) fat rather than just overall body weight.
The study identified specific thresholds: a BRI of 3.66 or higher and a WHtR of 0.52 or higher were associated with roughly double the odds of having hypertension compared to people below those thresholds. Each standard deviation increase in either measure was linked to approximately 50% higher odds of hypertension. Importantly, the relationship between these measures and hypertension was not a simple straight line — it was nonlinear, meaning risk accelerated at certain levels rather than increasing at a constant rate.
This research suggests that waist-based measures of body shape, particularly BRI and WHtR, may be more useful than traditional weight-based measures like BMI for identifying adults at elevated hypertension risk in large-scale health surveys. The identified cut-off values could potentially help in public health screening efforts, though the cross-sectional design of the study means it cannot confirm that higher abdominal obesity directly causes hypertension.
Wang C, Wu Y, Nie M, Duobujie, Xiao C, He W, et al.. (2026). Associations between various obesity-related indicators and hypertension: a national cross-sectional survey with 117,609 Chinese adults.. Frontiers in public health. https://doi.org/10.3389/fpubh.2026.1925175