**Background:** Type 2 diabetes mellitus (T2DM) and sarcopenia are both common in aging populations, and their interaction is of growing interest. Sarcopenia is characterized by age-related loss of skeletal muscle mass, low muscle strength, or low physical performance, increasing risks of falls, fractures, disability, and mortality. Obesity, particularly abdominal obesity, may influence sarcopenia risk, but the relationship between different obesity phenotypes and sarcopenia in T2DM patients remains controversial. This study aimed to investigate the association between obesity phenotypes (defined by BMI and waist circumference) and sarcopenia in hospitalized Chinese middle-aged and elderly patients with T2DM.
**Methods:** This cross-sectional study included 385 hospitalized patients (mean age not reported as overall mean; age range ≥50 years) with T2DM from October 2019 to May 2022 at the Department of Endocrinology, First Hospital of Qinhuangdao, Hebei Province, China. Exclusion criteria included acute diabetic complications, acute cardiovascular or cerebrovascular events, acute inflammation, malignancy, severe hepatic or renal dysfunction, and inability to perform exercise tests. Body composition was measured by Dual-energy X-ray absorptiometry (DXA). Appendicular skeletal muscle mass index (ASMI) was calculated as sum of lean mass of bilateral limbs divided by height squared. Sarcopenia was diagnosed per the Asian Working Group for Sarcopenia 2019 (AWGS 2019) consensus: low ASMI (<7.0 kg/m² in males; <5.4 kg/m² in females) plus either low handgrip strength (<28 kg in males; <18 kg in females) or low gait speed (<1.0 m/s). Patients were divided into four groups: Group A (normal BMI <24 kg/m² and normal WC: male <90 cm, female <85 cm; n=67), Group B (normal BMI, high WC; n=60), Group C (high BMI ≥24 kg/m², normal WC; n=33), and Group D (high BMI, high WC; n=225). Statistical analyses included two-way ANOVA, chi-squared tests, and logistic regression.
**Key Results:** The overall prevalence of sarcopenia was 32.2%, and abdominal obesity prevalence was 74.0%. The detection rate of low ASMI decreased progressively from Group A to Group D: 74.6% (Group A), 68.3% (Group B), 54.5% (Group C), and 51.6% (Group D) (χ²=14.243, P=0.003). Sarcopenia prevalence was 41.8% in Group A, 36.7% in Group B, 27.3% in Group C, and 28.9% in Group D (χ²=4.868, P=0.182). Logistic regression adjusting for gender, age, HbA1c, and diabetes duration showed that compared with Group A, the risk of low ASMI was reduced by 62.4% in Group C (OR=0.376, 95% CI: 0.149–0.950, P=0.039) and by 68.8% in Group D (OR=0.312, 95% CI: 0.165–0.593, P=0.000). Male sex increased the risk of low ASMI by 4.153-fold (OR=4.153, 95% CI: 2.623–6.576, P=0.000). In a separate logistic regression among overweight/obese participants (Groups C and D), male sex (OR=4.065, 95% CI: 2.246–7.356, P=0.000) and higher WC (OR=1.053, 95% CI: 1.004–1.104, P=0.033) were risk factors for low ASMI, while higher BMI was protective (OR=0.680, 95% CI: 0.574–0.804, P=0.000), reducing risk by 32% per unit BMI increase. Gait speed was significantly lower in abdominal obesity groups (Group B: 0.96±0.24 m/s; Group D: 1.00±0.22 m/s) compared with Group A (1.08±0.21 m/s) (P=0.013). No significant differences were found in grip strength among groups.
**Clinical Implications:** This study demonstrates that in hospitalized middle-aged and elderly Chinese patients with T2DM, the prevalence of sarcopenia (32.2%) and abdominal obesity (74.0%) is high. The paradoxical finding that higher BMI is protective against low muscle mass while abdominal obesity (high WC) increases risk highlights the importance of assessing obesity phenotype rather than BMI alone. Clinicians should monitor waist circumference in T2DM patients to identify those at higher risk for sarcopenia. Interventions targeting visceral fat reduction may help preserve muscle mass. The cross-sectional design precludes causal inference, and the inpatient population may have more severe disease than community-dwelling individuals, limiting generalizability. Future prospective studies in community settings are needed to confirm these findings.