Risk prediction model of frailty and its associated factors in older adults: a cross-sectional study in Anhui Province, China.

Background: In the context of aging in China, frailty has become a major public health challenge, placing an enormous economic burden on both society and families. Frailty can trigger serious adverse effects on the physical and mental health of older adults. It highlights the urgent requirement for addressing the issue of frailty among older adults. Accordingly, the present study was conducted to identify potential risk factors and develop a validated risk predictive model for frailty in older Chinese adults.

Methods: Following a cross-sectional design, the present study selected participants from Anhui Province, China, using convenience sampling. Eligible data were collected using a demographic questionnaire, the Fatigue, Resistance, Ambulation, Illnesses, & Loss of Weight (FRAIL) scale, the strength, assistance walking, rise from a chair, climb stairs, and falls (SARC-F) scale, the social FRAIL scale, and the short-form mini-nutritional assessment (MNA-SF). Furthermore, a one-way analysis of variance and a multivariate analysis were utilized to identify the optimal predictive factors of the model. The logistic regression model was used to explore frailty-associated factors in older Chinese adults. Finally, a nomogram was constructed to establish the predictive model, with the application of calibration curves to evaluate the accuracy of the nomogram. The area under the receiver operating characteristic (ROC) curve (AUC) and decision curve analysis (DCA) were used to evaluate the performance of prediction.

Results: Our final analysis incorporated 1,611 older Chinese adults who completed the questionnaire, with the incidence of frailty found in 491 (30.5%) cases. Multivariate logistic regression analysis showed that age, sarcopenia, malnutrition, social frailty, and hospitalization within the past 6 months were predictors of frailty. Consequently, the resultant nomogram demonstrated good consistency and accuracy. The AUC values of the model and the internal validation set were 0.86 (95%CI: 0.84-0.89) and 0.89 (95%CI: 0.85-0.92), respectively (both  > 0.05 via the Hosmer-Lemeshow test). In addition, the calibration curve showed significant agreement between the nomogram predictions and the observed values. ROC and DCA analyses revealed good predictive performance of the nomogram.

Conclusion: This study constructs a frailty risk predictive model with good consistency and predictive performance, facilitating an effective prediction of the onset of frailty among older Chinese adults. It may benefit the screening of high-risk populations and the implementation of early interventions clinically.