Association between body roundness index and incident stroke with different blood pressure status: A retrospective propensity score matched analysis of the CHARLS study

Abstract

Background and aim

Body roundness index (BRI), an innovative anthropometric measure assessing visceral fat, has demonstrated utility in predicting cardiometabolic risk. However, its association with stroke risk across blood-pressure strata remains unclear.

Methods and results

The sample comprised 12,316 CHARLS participants aged ≥45 years without prior stroke. The association between the BRI and incident stroke was evaluated using Cox proportional hazards models. To strengthen the validity of the findings, additional analyses were performed, including propensity score matching (PSM), subgroup analyses, and sensitivity tests. Furthermore, the discriminative capacity of BRI for predicting stroke events was assessed using receiver operating characteristic (ROC) curve analysis.

Increased stroke risk was significantly connected to a higher BRI. Following PSM, fully adjusted models indicated that a unit rise in log (BRI) was tied to a 19 % increase in stroke risk (HR = 1.79, 95 % CI: 1.37–2.34, P < 0.001). After stratification by blood pressure status, the association between BRI and stroke risk was most pronounced among individuals with prehypertension (HR = 2.60, 95 %CI: 1.49–4.54; P < 0.001) and those with hypertension (HR = 1.65, 95 %CI: 1.17–2.33; P = 0.004). By contrast, among participants with normal blood pressure (NBP), no statistically significant association was observed following PSM. The reliability of the findings was supported by subgroup and sensitivity analyses. The ROC analysis demonstrated that the BRI had moderate predictive accuracy for stroke, notably in individuals with NBP, with an area under the curve of 0.672.

Conclusions

Elevated BRI is independently associated with a greater risk of stroke, particularly in individuals with prehypertension or hypertension.