The serum creatinine-to–cystatin C ratio (Cre/CysC) and percent body fat (PBF) have each been linked to reduced muscle mass and impaired muscle function. However, their combined ability to predict structural and functional muscle deficits in type 2 diabetes mellitus (T2DM) remains insufficiently characterized. This study evaluated whether integrating Cre/CysC with PBF—and additionally with age—improves discrimination of low muscle mass, weak muscle strength, and slow gait speed.

We conducted a cross-sectional study of 93 adults (≥18 years) with T2DM recruited from the outpatient clinic of the University Medical Center Ho Chi Minh City, Campus 2 (UMC2). Body composition was measured by bioelectrical impedance analysis based on FINH 2014 criteria, including appendicular skeletal muscle mass (ASM), ASM/BMI, and PBF. Muscle function was assessed using handgrip strength (HGS) and gait speed (GS). Serum creatinine and cystatin C were used to calculate the Cre/CysC ratio. Two predictive models were built: (1) PBF + Cre/CysC and (2) PBF + Cre/CysC + age. Discriminative performance for low ASM/BMI, weak HGS, and slow GS was evaluated using ROC curves and AUC values.

Higher Cre/CysC values were associated with greater ASM/BMI, stronger HGS, and faster GS. For detecting low muscle mass, AUC increased from 0.82 (Model 1) to 0.85 (Model 2). For weak muscle strength, AUC improved from 0.76 to 0.81, and for slow gait speed, from 0.74 to 0.78 after adding age. Across all outcomes, the full model (PBF + Cre/CysC + age) consistently demonstrated the highest discriminatory ability.

A combined model using Cre/CysC and PBF provides strong predictive accuracy for identifying muscle impairment, and the addition of age further enhances performance. This minimal-biomarker approach may facilitate early recognition of muscle decline in adults with T2DM.