Related Factors And Risk Assessment Of Depression In Middle-aged And Elderly People

ObjectiveThis study aims to gain a deeper understanding of the current state of depression among middle-aged and elderly people in my country.Study the comprehensive effect of multiple factors on depression and the interaction effects of multiple factors.Establish a depression risk assessment model for middle-aged and elderly people to assess the risk of depression in middleaged and elderly people,and provide a reference for early identification,early prevention,and timely intervention of depression.Combined with the results of empirical analysis,the corresponding policy improvement suggestions are put forward.MethodsUsing the 2011 China Health and Elderly Tracking Survey(CHARLS)data,the effective sample size was 4965 cases,and the score of 10 or more on the CESD-10 scale was defined as depressive symptoms.First,the descriptive statistics method is used to analyze the current depression situation of middle-aged and old people.Then establish a multi-factor Logistic regression model to explore the influencing factors of depression in the elderly.Finally,Bayesian network learning is carried out on the risk factors,and a depression risk assessment model for middle-aged and elderly people is established to predict the depression status of the middle-aged and elderly people.Results1.Among the 4965 subjects,1750(35%)with depressive symptoms,3215(65%)without depressive symptoms;2470(49.7%)males and 2495(50.3%)females;the average age was 59.6±9.6( ± );3735(75.2%)cases in rural areas,1230(24.8%)cases in cities and towns;3064(61.7%)cases in primary school and below;4117(82.9%)cases with spouses,848 without spouses(17.1%)cases;2665(53.7%)cases willing to live with children,2300(46.3%)cases not willing to live with children;712(14.3%)cases with physical dysfunction,4253 with normal physical function(85.7%)cases;3366(67.8%)cases with chronic diseases and 1599(32.2%)cases without chronic diseases.2.The three indicators with the highest scores on the CESD-10 scale are "full hope for the future"(average score: 1.117),"very enjoyable"(average score: 1.015)and "bad sleep"(average score: 1.009);The three lowest indicators are "feel scared"(average score: 0.316),"feel unable to continue living"(average score: 0.354)and "feel lonely"(average score: 0.546).3.The results of multivariate logistic regression analysis showed: gender,urban and rural areas,education level,marital status,sleep time,whether to drink alcohol,self-assessment of health,overall life satisfaction,physical function,chronic disease,frequent physical pain,pension,low-income families,Family income and community economic status are significant factors influencing depression of middle-aged and elderly people,P<0.05.4.The Bayesian network model consists of a network structure with 16 network nodes and 24 directed edges and 16 conditional probability tables.Bayesian network model 10-fold cross-validation results,the accuracy rate is 0.715.200 cases of 2015 CHARLS survey data were selected for external verification,and the accuracy rate was 0.710.Gender,physical pain,selfassessment of health,urban and rural areas,education level,marital status and family income are the sensitive factors that affect the depression of middle-aged and elderly people.The two largest causes of depression for middle-aged and elderly people are: physical pain—>depression,physical pain—>health self-assessment—>depression.Conclusion1.The depression situation of the middle-aged and elderly people in my country is not optimistic.The feelings of depression in the middle-aged and elderly people with low future expectations,low mood,and poor sleep are more obvious than other indicators.2.Sociodemographic characteristics,subjective factors,and objective factors all have a significant impact on the occurrence of depression in middle-aged and elderly people.3.Applying the Bayesian network to the assessment of depression risk of middle-aged and elderly people can intuitively understand the relationship between factors influencing depression status,effectively predict individual depression risk,and analyze the largest cause chain of depression.It shows the advantages of Bayesian network in medical research and the feasibility of practice in other diseases.