Construction and evaluation of malnutrition risk prediction and scoring system in Parkinson's disease patients

Abstract

Objective To understand the clinical nutritional status and malnutrition risk factors of Parkinson's disease (PD) patients and construct their risk scoring system. Methods A total of 221 patients with PD admitted to our hospital from January 2015 to January 2019 were enrolled. The nutritional status of the patients was assessed by Mini Nutrition Assessment (MNA), and according to the MNA scores, these patients were divided into well-nourished groups (n=176) and malnourished group (including patients having malnutrition and malnutrition risk, n=45). Single factor analysis and Logistic regression analysis were used to analyze the risk factors; their scoring systems were constructed and Hosmer-Lemeshow goodness of fit test was used. Accuracy and discrimination of evaluation scoring systems were evaluated by area under receiver operating characteristic (ROC) curve. Results A total of 45 patients (20.36%) had malnutrition or risk of malnutrition. Multivariate regression analysis suggested that the course of disease≥9 years, motor symptoms (unified PD rating scale III scores≥54), Hoehn-Yahr (H-Y) staging≥4, daily levodopa equivalent dose≥600 mg, scores of non-motor symptom scale (NMSS) scores≥101, and quality of life (39-item PD questionnaire scores≥81) were independent risk factors for malnutrition in PD patients (P<0.05). The constructed risk scoring system scored from 0 to 30, and Hosmer-Lemeshow goodness of fit test were: χ2=6.259, P=0.618; the area under ROC curve was 0.830 (95%CI: 0.801-0.860); when the predicted score was 14, Yoden index was the largest (0.517), and the sensitivity and specificity were 75.70% and 76.00%, respectively. Conclusions PD patients have a high risk of malnutrition or malnutrition. The risk scoring system based on risk factors has high predictive ability and discriminating ability, and can be used as an important tool for malnutrition risk assessment in PD patients. Key words: Parkinson's disease; Malnutrition; Risk factor; Prediction scoring system