Nomogram models for predicting sarcopenia in elderly Asian patients with type 2 diabetes.

BACKGROUND

Sarcopenia is a prevalent but underrecognized complication in elderly patients with Type 2 Diabetes Mellitus (T2DM). Its complex etiology limits early diagnosis and intervention. This study developed and internally validated a nomogram for individualized sarcopenia risk assessment in this population.

METHODS

A cross-sectional study was conducted involving 300 elderly patients diagnosed with T2DM from two tertiary medical centers. Sarcopenia was identified based on the Asian Working Group for Sarcopenia (AWGS) criteria, incorporating both dual-energy X-Ray Absorptiometry (DXA) measurements and handgrip strength assessments. Potential risk predictors were initially filtered using the Least Absolute Shrinkage and Selection Operator (LASSO) method, followed by multivariable logistic regression to ascertain independent factors. A nomogram model was then established integrating these variables. Model performance was assessed through Receiver Operating Characteristic (ROC) curve analysis, calibration plots, and Decision Curve Analysis (DCA) for clinical applicability.

RESULTS

Sarcopenia was present in 22 % of participants. Independent predictors included age, sex, glycosylated Hemoglobin (HbA1c), serum albumin, thyroid dysfunction, and physical activity level. The nomogram demonstrated strong predictive accuracy, with an area under the ROC curve of 0.891 in the development cohort and 0.868 in the validation cohort. Calibration and decision curve analyses confirmed its reliability and clinical utility.

CONCLUSIONS

The proposed nomogram enables early identification of sarcopenia risk in elderly patients with T2DM, supporting personalized intervention strategies. Prospective validation in diverse populations is warranted to enhance its generalizability.