# The Clínica Universidad de Navarra-Body Adiposity Estimator index is a reliable tool for prediabetes: a multi-center retrospective cohort study

## TL;DR

The CUN-BAE index is a reliable tool for predicting prediabetes risk, outperforming traditional metrics like BMI.

## Contribution

This study demonstrates the CUN-BAE index's independent and nonlinear association with prediabetes risk across diverse populations.

## Key findings

- CUN-BAE is independently and nonlinearly positively associated with prediabetes risk.
- CUN-BAE outperforms BMI, TG/HDL-c, and TyG in predicting prediabetes risk.
- Subgroup analyses reveal gender-specific risk factors interacting with CUN-BAE.

## Abstract

Given the close link between body fat and insulin resistance, our study aimed to evaluate the association and predictive value of the Clínica Universidad de Navarra Body Adiposity Estimator (CUN-BAE) in identifying the risk of prediabetes.

This multi-center retrospective cohort study included 112,708 participants from 32 regions across 11 cities in China. Multivariable Cox regression and restricted cubic spline (RCS) analyses were used to assess the association between CUN-BAE and prediabetes risk. Predictive effectiveness was evaluated using receiver operating characteristic (ROC) curves.

An independent association of CUN-BAE with prediabetes was significantly shown after adjusting for important covariates. CUN-BAE is nonlinearly positively correlated with prediabetes. In the male subgroup analysis, higher triglyceride levels or hypertension in combination with CUN-BAE significantly increased the incidence of prediabetes (interaction p < 0.05 for both). In the female subgroup analysis, middle age, BMI > 28, higher triglyceride levels, or hypertension in combination with CUN-BAE significantly increased the incidence of prediabetes (interaction p < 0.05 for all). ROC analysis demonstrated that CUN-BAE performed better in predicting risk than BMI, TG/HDL-c and TyG.

The CUN-BAE index was independently and nonlinearly positively associated with an increased risk of prediabetes and exhibited high predictive accuracy.