BRI showed a nonlinear, model-dependent association with prevalent self-reported stroke, and frailty provided additional cross-sectional information, supporting prospective evaluation of BRI and frailty as complementary markers in stroke-related assessment.
Higher body roundness index (BRI) was associated with prevalent self-reported stroke in the primary adjusted model.
The association between BRI and prevalent self-reported stroke was nonlinear, showing increasing odds across lower-to-middle BRI values followed by attenuation at higher values.
The BRI-stroke association was highly model-dependent, changing direction when the frailty index (FI) was included or excluded.
The frailty index (FI) was strongly associated with prevalent self-reported stroke, independent of BRI.
Adding BRI to age and sex modestly improved stroke discrimination, while adding the frailty index produced a substantially larger improvement.
The study sample consisted of 36,324 adults drawn from NHANES data spanning 1999 to 2023.
This research examined whether body roundness index (BRI) — a measure of body shape that estimates how 'round' a person is based on height and waist circumference — is associated with having had a stroke, and whether physical frailty plays a role in that relationship. Using data from over 36,000 American adults collected between 1999 and 2023, the researchers found that BRI was indeed associated with a history of stroke, but the relationship was curved rather than straight: stroke odds increased as BRI rose from low to moderate values, but leveled off or attenuated at higher BRI values. A key turning point in the curve was estimated at a BRI value of about 6.34. A particularly striking finding was how much the results depended on whether frailty was accounted for in the analysis. When frailty was not included, BRI showed essentially no association with stroke. When frailty was included, BRI's association actually reversed direction. Meanwhile, frailty itself was very strongly linked to stroke history — people with higher frailty scores had more than twice the odds of having had a stroke for every 0.1-unit increase in the frailty index. In terms of identifying who has had a stroke, adding frailty to the model dramatically improved accuracy (from an AUC of 0.769 to 0.868), while adding BRI alone provided only a small improvement. This research suggests that body shape (as measured by BRI) and frailty may both be relevant to understanding stroke risk, but that frailty is a much more informative factor. The strong dependency of BRI's association on whether frailty is accounted for indicates that these two measures are closely intertwined. The authors call for future studies that follow people over time to better understand whether BRI and frailty together can help predict stroke before it occurs, rather than simply identifying patterns in people who have already experienced one.
Hu Z, Zhuang J, Wang J, Jin X, Fang L, Hu J. (2026). Body roundness index, frailty, and prevalent self-reported stroke: An exploratory cross-sectional decomposition analysis.. Annals of the Academy of Medicine, Singapore. https://doi.org/10.47102/annals-acadmedsg.2025411