We read with great interest the article entitled"Negative Association Between Body Roundness Index and Bone Mineral Density: Insights from NHANES"by Ding et al. (1). Using nationally representative data from the National Health and Nutrition Examination Survey (NHANES), the authors applied weighted regression models and smooth curve-fitting techniques to examine the association between Body Roundness Index (BRI) and bone mineral density (BMD). However, the significant reversal in the direction of association during model adjustment raises about model dependency and casts doubt on robustness of the reported 21 negative correlation.The authors constructed three weighted regression models to examine the association between BRI and BMD. 24 Notably, the estimated association exhibited a marked direction reversal across sequential adjustments, 25 shifting from a non-significant relationship in the unadjusted model to a positive association after partial 26 adjustment, and ultimately to a statistically significant negative association in the fully adjusted model. Such 27 direction switching raises concerns regarding model dependency and statistical instability, suggesting that the 28 reported inverse association may be contingent upon covariate structure rather than reflecting a stable 29 independent effect of BRI.
Guo et al. (Thu,) studied this question.