Association between obesity indices, insulin resistance markers, and osteoarthritis in middle-aged and elderly Chinese adults

TL;DR

This study finds that obesity and insulin resistance are linked to a higher risk of osteoarthritis in middle-aged and elderly Chinese adults, with TyG-WHtR being the strongest predictor.

Contribution

The study introduces TyG-WHtR as a novel and effective predictor of osteoarthritis risk in a Chinese population.

Findings

All obesity and insulin resistance indices were positively associated with osteoarthritis risk.

TyG-WHtR showed the highest predictive ability for OA with an AUC of 0.680.

Each unit increase in TyG-WHtR was linked to a 20% higher OA risk in a dose–response manner.

Abstract

Previous studies have indicated an association between osteoarthritis (OA), obesity, and insulin resistance (IR). However, current literature lacks sufficient clinical data to fully elucidate the relationship between obesity indices, insulin resistance surrogates (IR surrogates), and OA in China's middle-aged and elderly population. This study aims to investigate the correlation between obesity indices [body fat percentage (BFP), lipid accumulation product (LAP), body mass index (BMI), waist-to-height ratio (WHtR)], IR surrogates [triglyceride-glucose (TyG) index and its derivatives: TyG with waist circumference (TyG-WC), TyG-BMI, TyG-WHtR, and OA risk, and evaluate the diagnostic efficacy of these indices for OA.

This study utilized data from the China Health and Retirement Longitudinal Study (CHARLS). Multivariable logistic regression and Cox proportional hazards models were employed, alongside Receiver Operating Characteristic (ROC) curves, restricted cubic splines, and subgroup analyses, to assess the associations between obesity indicators, IR surrogates, and the risk of OA in middle-aged and older adults.

A multivariable logistic regression analysis was conducted using data from 10,457 participants, of whom 3,667 were diagnosed with OA. In fully adjusted models, all indices as continuous variables were positively associated with OA risk (all p < 0.05): BFP (95% CI: 1.02–1.04), LAP (95% CI: 1.04–1.15), BMI (95% CI: 1.02–1.05), WHtR (95% CI: 1.10–1.21), TyG (95% CI: 1.02–1.20), TyG-WC (95% CI: 1.06–1.18), TyG-BMI (95% CI: 1.10–1.22), and TyG-WHtR (95% CI: 1.14–1.32). ROC analysis indicated TyG-WHtR had the greatest predictive ability for OA risk (AUC = 0.680). A multivariable Cox regression analysis of TyG-WHtR in 5,718 participants, among whom 1,827 developed OA during a median follow-up of 108 months, showed each one-unit increase in TyG-WHtR was associated with a 20% higher risk of OA (95% CI: 1.11–1.31). Trend tests revealed a significant dose–response relationship (p < 0.05).

Obesity-related indicators and IR surrogates are significantly associated with OA risk. Among these, TyG-WHtR demonstrates the strongest predictive performance, suggesting its potential as an early screening tool for OA. This study highlights obesity and IR as modifiable risk factors, providing a basis for the early prevention and control of OA.

Figures