Association Between Sarcopenia Index and Central Obesity: Age- and Gender-Specific Patterns and Clinical Implications

Central obesity is a major contributor to cardiometabolic diseases. The sarcopenia index (SI), derived from serum creatinine and cystatin C, is a biomarker reflecting muscle mass, but its relationship with obesity phenotypes across age and sex remains unclear.

To examine the association between SI and central versus peripheral obesity, determine whether SI is independently associated with central obesity, and identify optimal SI cut-off values in a large Chinese population.

This retrospective cross-sectional study analyzed 10,054 adults undergoing health examinations at Xiamen Chang Gung Hospital. Anthropometric data, biochemical profiles, and body composition by bioelectrical impedance analysis were collected. Participants were categorized as non-obese, peripheral obese, or central obese based on body fat percentage and visceral fat degree. Logistic regression assessed associations between SI tertiles and central obesity stratified by age and sex. Receiver operating characteristic (ROC) curves determined optimal SI cut-off values.

Central obesity was present in 24.4% of participants and correlated with adverse cardiometabolic profiles. Individuals with central obesity exhibited lower SI levels, most notably in men <50 years and women ≥50 years. After multivariable adjustment, the lowest SI tertile was independently associated with central obesity in younger men (adjusted OR 1.27; 95% CI 1.06-1.52) and older women (adjusted OR 3.03; 95% CI 2.04-4.50). SI demonstrated moderate discriminatory ability in women (AUC 0.687; optimal cut-off 79.52) but limited performance in men (AUC 0.534; cut-off 96.60).

Lower SI is independently associated with a higher prevalence of central obesity, particularly among younger men and older women. SI may provide a simple and practical supplementary marker for identifying individuals at risk of central adiposity, especially in women. Prospective studies are needed to validate its predictive utility.