Association between CVAI-defined body composition phenotypes and prediabetes outcomes in Chinese adults undergoing...

Prediabetes presents a critical opportunity to prevent type 2 diabetes and its complications. Both visceral adiposity and low skeletal muscle mass are closely linked to insulin resistance; however, it remains unclear whether body composition phenotypes-defined by relative muscle mass and the Chinese visceral adiposity index (CVAI), which is a validated measure of visceral fat accumulation in Chinese adults-are differentially associated with specific prediabetes outcomes. Therefore, we examined the associations between CVAI-defined body composition phenotypes and impaired fasting glucose (IFG), elevated HbA1c levels, and composite prediabetes and compared the discriminatory performance of the CVAI with that of BMI and the waist-to-height ratio (WHtR).

Data obtained from 3,905 adults who attended a health check-up center in 2024 were analyzed. Participants were classified into four body composition phenotypes: Normal, low relative muscle mass alone, visceral obesity alone, and combined low relative muscle mass with visceral obesity. Associations with each prediabetes outcome were estimated using logistic regression to obtain odds ratios (ORs) and 95% confidence intervals (CIs) across three models: Model 1, unadjusted; Model 2, adjusted for age and sex; and Model 3, further adjusted for systolic blood pressure, diastolic blood pressure, low-density lipoprotein cholesterol (LDL-C), alanine aminotransferase (ALT), and estimated glomerular filtration rate (eGFR). The discriminatory performance of the CVAI, BMI, and waist-to-height ratio was evaluated using receiver operating characteristic curves (AUCs), with between-index comparisons conducted using the DeLong test.

Among the 3,905 adults, the phenotype distribution was as follows: 68.0% had a normal phenotype, 6.9% had low relative muscle mass alone, 11.9% had visceral obesity alone, and 13.2% had a combination of low relative muscle mass and visceral obesity. After adjusting for age and sex, the phenotype combining low relative muscle mass with visceral obesity showed the strongest associations with IFG, elevated HbA1c levels, and composite prediabetes (ORs 2.42-2.84). In contrast, obesity alone remained associated with elevated HbA1c levels and composite prediabetes, and low relative muscle mass alone was not independently associated. These findings remained materially unchanged in the extended confounder-adjusted model.

The phenotype characterized by low relative muscle mass combined with high CVAI-defined visceral adiposity was most strongly associated with impaired fasting glucose, elevated HbA1c levels, and composite prediabetes, whereas low relative muscle mass alone showed no independent association after adjustment. The CVAI also demonstrated better discrimination of abnormal glycemia compared to BMI and the waist-to-height ratio.