Abstract
OBJECTIVE: To identify distinct latent frailty profiles using latent profile analysis (LPA) in patients undergoing radiotherapy for head and neck cancer (HNC) and to examine the factors associated with profile membership. DESIGN: A cross-sectional study. SETTING: This research used data acquired from a major tertiary referral hospital in China. PARTICIPANTS: This study recruited 391 HNC patients receiving radiotherapy. OUTCOME MEASURES: Validated instruments included a demographic questionnaire, Charlson Comorbidity Index (CCI), Pittsburgh Sleep Quality Index, FRAIL Scale, Hospital Anxiety and Depression Scale and Perceived Social Support Scale. Profile membership associations were assessed using χ(2) tests, univariate analysis and multivariable logistic regression. RESULTS: The frailty status of patients can be divided into three different categories: (1) robust group (23.0%), (2) prefrail group (49.6%) and (3) frail group (27.4%). Frailty demonstrated independent associations with nine clinical parameters in adjusted regression models: radiotherapy session frequency, social support, age, CCI score, educational attainment, metastasis, nutritional risk, radiation-induced injuries and serum albumin levels (p<0.05). CONCLUSION: The distinct frailty profiles identified by LPA can inform the future development of targeted care protocols for specific subgroups (eg, the frail group), with a focus on key predictors such as age and nutritional risk.