关键词: 衰弱, 炎性因子, 列线图, 影响因素, Logistic回归分析

Abstract: Background Frailty is a common geriatric syndrome, and there is currently no unified gold standard for assessments of this condition, which mostly depends on a range of measures. Chronic inflammation as one of the pathophysiologic mechanisms of frailty and the simple availability of inflammatory factors in the blood routine, there are fewer studies on the correlation between them and frailty. Objective The goals of this study are to build a columnar graphical model of the risk of acquiring frailty, assess the factors influencing frailty, and look into the relationship between routine blood inflammatory indicators and frailty in older adults receiving physical examination. Methods In August 2020 to December 2022, 554 elderly people who had health examinations participated in a cross-sectional study. The influencing elements of frailty were analyzed using logistic regression, column-line plots were created based on the influencing factors, and the model was internally validated using bootstrap. The subjects' working curve, the clinical decision curve graph, and the Homster-Lemeshow goodness-of-fit test in conjunction with the calibration curve graph were used to assess the model's performance, respectively. Results A total of 554 older adults were included, the mean age was 74.4±9.5 years, and 213 (38.4%) were frail. In multivariate logistic regression analysis we identified several variables independently associated with the risk of frailty: ACCI (OR=1.42, 95% CI 1.21-1.66), MNA.SF (OR=0.71, 95% CI 0.61-0.83), HRR (OR=0.44, 95% CI 0.23-0.86), and polypharmacy (OR=0.54,95% CI 0.36-0.81). The column line diagram works by creating a multifactor regression model that has a 0.721 area under the ROC curve. The column-line diagram model had an excellent fit (P<0.05) after internal validation using the Bootstrap resampling approach; the calibration curve diagrams demonstrated a high match between the anticipated probability and the actual probability of occurrence. The model's strong clinical utility was demonstrated by the clinical decision plot. Conclusion Frailty was influenced by age, comorbidities, polypharmacy, malnourishment, and HRR. The developed prediction model can offer guidelines for early frailty screening and has strong distinction, consistency, and clinical utility.

Key words: Frailty, Inflammatory factors, Column-line diagram, Influencing factors, Logistic regression analysis