A Nomogram for Individualized Frailty Identification in Older Inpatients with Schizophrenia: Development and Internal Validation

Objective: To investigate the prevalence and associated factors of frailty among older inpatients with schizophrenia, and to develop and internally validate a nomogram for individualized identification of current frailty. Methods: Between May and September 2025, we enrolled 279 older inpatients with schizophrenia (aged ≥ 60 years) from a tertiary hospital in Guangzhou using convenience sampling. Data included electronic medical records, patient interviews, and the Fried frailty phenotype assessment. Variables with univariate P < 0.1 or pre-specified clinical importance were refined by least absolute shrinkage and selection operator (LASSO) (10-fold cross-validation, lambda.1se) and binary logistic regression. A nomogram was built and internally validated using bootstrap (1,000 iterations). Performance was evaluated by the area under the receiver operating characteristic curve (AUC; DeLong’s method), calibration (Hosmer–Lemeshow test), and decision curve analysis (DCA). The study was approved by the Institutional Review Board of The Affiliated Brain Hospital of Guangzhou Medical University (Approval No: 2025– 034). Written informed consent was obtained from all participants or their legal guardians for those with impaired capacity. Results: Frailty prevalence was 38.0% (≥ 3 Fried criteria). Five clinical factors (represented by six variables) were retained: age (OR = 1.070, 95% CI: 1.018– 1.124), fall history (OR = 2.480, 95% CI: 1.027– 5.987), walking aid (OR = 4.021, 95% CI: 0.996– 16.228; borderline significance), sensory impairment (single: OR = 2.145, 95% CI: 1.051– 4.377; dual: OR = 3.470, 95% CI: 1.562– 7.708), and ADL (OR = 0.967, 95% CI: 0.936– 0.999). The model showed acceptable discrimination (AUC 0.767, 95% CI: 0.708– 0.826), good calibration (Hosmer–Lemeshow P = 0.552; Brier score 0.182; calibration slope 0.967). DCA showed positive net benefit at thresholds 0.1– 0.4. Conclusion: The nomogram showed acceptable discrimination and good calibration for identifying frailty in older inpatients with schizophrenia. Prospective multicenter external validation is needed.