Abstract
BACKGROUND AND OBJECTIVES: Patients with schizophrenia face an elevated risk of sarcopenic obesity (SO) due to antipsychotic-induced metabolic dysfunction, physical inactivity, and nutritional deficiencies. Although recent studies suggest an association between SO and cognitive decline, its specific impact on cognitive function in schizophrenia remains to be fully elucidated. This study aimed to evaluate the diagnostic concordance between the European Society for Clinical Nutrition and Metabolism/European Association for the Study of Obesity (SOESPEN) criteria and its modified version (SOESPEN-M), and to examine their respective associations with cognitive function in inpatients with stable schizophrenia. METHODS: In this cross-sectional analysis, 228 adults with stable schizophrenia were recruited. SO was diagnosed using two definitions: SOESPEN (excess adiposity, low muscle mass-to-weight ratio, and reduced handgrip strength) and SOESPEN-M (BMI-adjusted muscle mass threshold). Cognitive function was assessed using the Montreal Cognitive Assessment-Chinese version (MoCA-C). Multivariate linear regression models were employed to evaluate associations between SO and MoCA-C scores, adjusting for relevant demographic, clinical, and comorbidity-related variables. RESULTS: SO prevalence was 17.1% under both diagnostic criteria, with moderate inter-criteria agreement (κ = 0.660). Sex-stratified analyses revealed divergent diagnostic trends: in males, SO prevalence increased from 15.9% (SOESPEN) to 22.5% (SOESPEN-M; κ = 0.698); in females, prevalence decreased from 18.9% to 8.9% (κ = 0.590). Across both criteria, SO groups demonstrated significantly lower MoCA-C scores (males: 16 vs 20, p = 0.045 for SOESPEN; 13 vs 21, p < 0.001 for SOESPEN-M; females: 11 vs 17, p = 0.009 for SOESPEN; 10.5 vs 17, p = 0.036 for SOESPEN-M). Multivariate analysis confirmed that SOESPEN-M-defined SO was independently associated with lower MoCA-C scores in males (β = -2.71, 95% CI: -5.08 to -0.33, p = 0.027). CONCLUSION: Our results demonstrate that SO defined by SOESPEN-M criteria is significantly associated with cognitive impairment in male inpatients with stable schizophrenia.