# Association between body roundness index trajectories and the incidence of diabetes mellitus: a perspective from the China health and retirement longitudinal study

## TL;DR

This study finds that people with consistently high body roundness index (BRI) over time are at higher risk of developing diabetes, suggesting long-term monitoring of body shape could help identify diabetes risk early.

## Contribution

The study introduces longitudinal body roundness index (BRI) trajectory modeling as a novel method for diabetes risk assessment.

## Key findings

- Participants with high-stable BRI had a 2.63 times higher risk of developing diabetes compared to those with low-stable BRI.
- Longitudinal BRI trajectories improved diabetes risk reclassification and discrimination beyond single-time-point measures.
- BRI trajectories showed comparable predictive power to BMI and waist circumference for diabetes risk.

## Abstract

To investigate the associations between longitudinal body roundness index (BRI) trajectories and the risk of incident diabetes mellitus (DM) using data from the China Health and Retirement Longitudinal Study (CHARLS).

Group-based trajectory modeling (GBTM) identified distinct BRI trajectories (Waves 1–3, 2011–2016). Their associations with DM incidence (Wave 4, 2017–2018) were assessed using multivariate Cox models. The predictive performance of a single baseline BRI was compared with body mass index (BMI) and waist circumference (WC) via receiver operating characteristic (ROC) analysis. Net reclassification improvement (NRI) and integrated discrimination improvement (IDI) evaluated the incremental value of adding BRI trajectories to a conventional risk model. Subgroup and sensitivity analyses, including a landmark approach, assessed robustness.

Among 4,150 participants, 103 developed DM. Three stable BRI trajectories were identified: low-stable (49.0%), moderate-stable (41.3%), and high-stable (9.7%). Compared with the low-stable group, the high-stable group had a significantly increased DM risk with a fully-adjusted hazard ratio (HR) of 2.63 (95% confidence interval [CI]: 1.41–4.91). A single baseline BRI showed comparable discrimination to BMI and WC (AUC ≈ 0.63). Longitudinal trajectories of BRI, BMI, and WC all identified high-stable subgroups with elevated risk (HRs: BRI = 2.63, BMI = 2.16, WC = 2.31), with overlapping confidence intervals. However, adding BRI trajectories to a conventional model significantly improved risk reclassification (NRI = 10.76%, 95% CI: 2.40–19.47) and discrimination (IDI = 0.27%, 95% CI: 0.03–0.52). Results were consistent across subgroups and sensitivity analyses.

Sustained high BRI exposure, captured by longitudinal trajectory modeling, is independently associated with increased DM risk. While BRI trajectories were not statistically superior to BMI or WC trajectories, the longitudinal framework itself adds value over single-time-point assessments by more robustly identifying individuals with persistent high adiposity-related risk, highlighting the utility of monitoring long-term body shape stability for early risk stratification.

The online version contains supplementary material available at 10.1186/s12944-025-02840-y.