Study On The Correlation Between Fatigue And Quality Of Life In Patients With Chronic Pancreatitis

Objectives1.To Investigate the status of fatigue among patients with chronic pancreatitis(CP)in China and analyze the main influencing factors for fatigue.2.To explore the relationship between fatigue,anxiety and depression and quality of life(QoL)in patients with CP.3.To construct predictive models using machine learning for patients with CP and explore the important predictors.Methods1.Patients with CP who were hospitalized in the Department of Gastroenterology of the First Affiliated Hospital of Naval Military Medical University in Shanghai from August 2022 to January 2023 were selected using the convenience sampling method.2.The self-designed "Quality of Life Questionnaire for Chinese CP Patients",including General Information Questionnaire,Multidimensional Fatigue Inventory-20(MFI-20),Pittsburgh Sleep Quality Index(PSQI),the Hospital Anxiety and Depression Scale(HADS),the Brief Pain Inventory(BPI)and the European The European Organization for Research and Treatment of Cancer Quality of Life Questionnaire(EORCT QLQ-C30)was used to assess anxiety,depression,fatigue,sleep quality,pain and QoL in Chinese CP patients.3.The SPSS 26.0 software,PROCESS v4.1 plug-in were used for statistical analysis.Statistical analysis included:descriptive statistics,independent sample t-test,Mann-Whitney U-test,Chi-square test,binary logistic regression analysis,correlation analysis and mediation effect test.Machine learning predictive models were developed using Python 3.8.5 and the Scikit-learn package 0.23.2.All data were randomly divided into a training set and a test set in the ratio of 7:3.The training set was used to construct logistic regression(LR),random forest(RF)and extreme gradient boosting(XGBoost)models,and the test set was used for internal validation of the models.AUC,accuracy(ACC),sensitivity(SEN),specificity(SPE),positive prediction rate(PPV),negative prediction rate(NPV)and F1 values were used to evaluate the performance of the models.Feature importance scores were used to evaluate the predictive value of the included predictors.Results1.Status and influencing factors for fatigue among patients with CP in China.A total of 262 patients with CP were included in this study,of whom 36.64%(96/262)had severe fatigue symptoms.The MFI-20 score of patients with CP was(54.02±12.59),including general fatigue(11.98±2.89),physical fatigue(11.56±3.32),mental fatigue(9.71 ±3.76),reduced activity(11.20±3.22)and reduced motivation(9.57±2.92).The scores for general fatigue,physical fatigue and reduced activity were significantly higher than those for psychological fatigue and reduced motivation.The results of univariate analysis showed that there is a statistically significant difference(P<0.05)in gender,age,work status,monthly household income,duration of illness,history of endoscopic treatment,history of pancreatic surgery,steatorrhea,history of smoking,history of drinking,history of diabetes and pain pattern among patients with CP.Binary logistic regression analysis showed that history of endoscopic treatment(OR=0.432,95%CI:0.202-0.924),steatorrhea(OR=3.637,95%CI:1.607-8.232),history of smoking(OR=4.912,95%CI:1.388-17.381),sleep disturbance(OR=5.523,95%CI:2.434-12.532),anxiety(OR=2.442,95%CI:1.011-5.897)and depression(OR=4.935,95%CI:2.143-11.364)were influential factors of fatigue for patients with CP(P<0.05).2.The relationship between fatigue,anxiety,and depression and QoL in patients with CP.CP patients with fatigue had significantly lower level of QoL.The score of EORCT QLQC30 global health status(42.06±17.56 vs.68.48±17.91,P<0.001)and functional scores:physical functioning(79.63±14.65 vs.91.75±9.23,P<0.001),role functioning(78.50±26.40 vs.92.92± 13.32,P<0.001),emotional functioning(61.04±21.47 vs.79.47± 18.36,P<0.001),cognitive functioning(68.38±20.47 vs.85.75±14.90,P<0.001),and social functioning(52.96±28.02 vs.77.93±22.25,P<0.001)were significantly lower;symptom scores:fatigue(47.56± 18.30 vs.23.71±12.26,P<0.001),nausea and vomiting(8.57±18.65 vs.4.19±10.43,P=0.026),pain(36.92±26.60 vs..20.48± 19.58,P<0.001),dyspnea(21.18±22.15 vs.8.93± 16.02,P<0.001),insomnia(41.12±30.23 vs.19.74±23.07,P<0.001),appetite loss(20.25±28.50 vs.9.68± 17.82,P=0.001),constipation(26.48±29.58 vs.16.57±22.17,P=0.006),diarrhea(24.61±29.44 vs.9.31±15.81,P<0.001),and financial difficulties(44.86±34.60 vs.24.21±27.55,P<0.001)were significantly higher.Correlation analysis showed that the total score of MFI-20 in patients with CP was significantly negatively correlated with EORCT QLQ-C30 general health status scores and functional scores(r=-0.667,P<0.001)and significantly positively correlated with anxiety(r=0.563,P<0.001)and depression(r=0.643,P<0.001)scores.The results of the mediation analysis showed that fatigue could have an indirect effect on patients,QoL through anxiety and depression,with mediating effects of 21.24%and 16.90%,respectively.3.Construction of machine learning models for fatigue among patients with CP.In this part of the study,a total of 57 variables from 96 CP patients with fatigue and 166 CP patients without fatigue were included for the construction of predictive model.27 variables with statistically significant differences were screened by univariate analysis,and LR,RF and XGBoost models were developed based on these 27 predictors.In the test set,the AUCs for LR,RF and XGBoost were 0.892,0.912 and 0.874,respectively,and the ACCs were 0.810,0.848 and 0.785,respectively.The results of relative feature importance showed that LR,RF and XGBoost models all suggested that anxiety,depression,sleep disturbance,steatorrhea,history of endoscopic treatment,HCT,RBC,GLU and chronic pain were significant predictors for fatigue in patients with CP.Conclusions1.Fatigue is a common symptom in Chinese CP patients.History of endoscopic treatment,steatorrhea,history of smoking,sleep disturbance,anxiety and depression are influencing factors for fatigue.Clinical and nursing staff should intervene early to address these factors:improve patients’ steatorrhea symptoms through dietary guidance and intensive pancreatic enzyme replacement therapy;persuade patients to adopt good lifestyle habits and quit smoking and drinking;and improve patients’ sleep disturbances and depression symptoms through psychological guidance and pharmacological interventions to reduce fatigue levels in patients with CP.2.Fatigue is significantly associated with QoL among CP patients.The higher the level of fatigue,the worse the QoL and the more severe the disease-related symptoms.In addition,fatigue can have an indirect impact on patients’ QoL through anxiety and depression.This finding suggests that fatigue and anxiety and depression may be a new way of affecting QoL in patients with CP,in addition to common disease symptoms such as pain and pancreatic exocrine insufficiency.Clinical staff should pay attention to patients’ fatigue and psychological comorbidities and adopt various means to reduce patients’ fatigue and anxiety and depression levels to improve their QoL.3.Predictive models based on machine learning have better predictive performance in predicting CP patients’ fatigue.Anxiety,depression,sleep disturbance,steatorrhea,history of endoscopic treatment,HCT,RBC,GLU and chronic pain are significant predictors for fatigue in patients with CP.This suggests that patients’ nutritional status and related indicators may be important factors influencing their fatigue levels and may provide help in predicting fatigue in patients with CP.