Study On The Logistic Regression Model Of Fall Risk In The Elderl

Objective:1.To determine the main risk factors of fall in the elderly,and to provide theoretical basis and scientific basis for effective research on the risk factors and risk level of fall in the elderly in China.2.Develop a fall risk assessment tool for the elderly based on Logistic regression model and evaluate the predictive effectiveness.3.Design the elderly fall risk assessment questionnaire to investigate the fall risk status and fall risk factors of the elderly in Beijing,and provide a basis for early preventive intervention.Methods:1.Based on systematic retrieval of Meta-analysis literature on elderly fall risk factors,the main risk factors for elderly fall were identified.Taking multi-disciplinary experts’opinions as reference and combining with China’s national conditions,the risk factors and relative risk of falling in the elderly were selected to be included in the model,and the risk assessment model of falling in the elderly in China was constructed by combining Logistic regression model2.Using a case-control study,the area under the ROC curve,sensitivity and specificity are the main evaluation indicators to evaluate the predictive power of the logistic regression model for fall risk in the elderly.3.Through a cross-sectional study,the main influencing factors of fall among the elderly in Beijing were investigated.Results:1.According to the search results of evidence-based medicine literature,the main risk factors for falls in the elderly include age,gender,fall history,nutritional status,living alone,exercise,use of walking aids,drinking,inappropriate shoes,wandering,fear of falling,self-perception of health,fatigue,frailty,Abnormal gait,abnormal balance,impaired mobility,self-care,dizziness/vertigo,visual impairment,hearing impairment,cognitive impairment,urinary/fecal incontinence,pain,sleep disturbance,muscle weakness,disability,sensory disturbance,sarcopenia,rheumatism/joint Inflammation,hypertension.History of stroke,dementia,diabetes,orthostatic hypotension,arrhythmia,Parkinson diseases,anemia,depression,foot diseases,eye diseases,cancer,antipsvchotics.antidepressants,sedatives/hypnotics,Antiepileptics,opioids,loop diuretics,cardiac glycosides,statins,hypoglycemic drugs,laxatives,PPI(proton pump inhibitor),etc.2.The constructed model is named the Chinese Fall Risk Assessment model for the Elderly(C-FRA model).The model included 32 indicators,namely age(1.43),history of falls(3.86),exercise(0.85),use of walking aids(2.39),frailty(1.84).abnormal gait(2.94).abnormal balance(2.26),Dizziness/vertigo(1.36),sleep disturbance(1.33),visual impairment(1.49),hearing impairment(1.37),cognitive impairment(1.32),urinaiy/fecal incontinence(1.73),lower extremity arthritis(1.80),diabetes(1.27),History of stroke(1.47),dementia(1.96),orthostatic hypotension(1.25),arrhythmia(1.42),Parkinson diseases(2.19),anemia(1.47),depression(1.64),foot problems(1.95),Antipsychotics(1.54),antidepressants(1.57),sedative/hypnotics(1.42),antiepileptics(1.55),opioids(1.60),loop diuretics(1.36),cardiac glycosides(1.60),hypoglycemic drugs(1.52),laxatives(1.43).3.The constructed fall risk assessment model for the elderly in C hina is as follows:Logit(P)=a+0.36x1+1.35x2-0.16x3+0.87x4+0.61x5+1.08x6+0.82x7+0.31x8+0.29x9+0.40x10+0.31x11+0.28x12+0.55x13+0.59x14+0.24x15+0.39x16+0.67x17+0.22x18+0.35x19+0.78x20+0.3 9x21+0.49x22+0.67x23+0.43x24+0.45x25+0.3 5x26+0.44x27+0.47x28+0.31x29+0.47x30+0.42x31+0.36x324.It has been verified that the area under the curve of the model in this study is 0.958(95%CI:0.914-0.983)and has good predictive performance.The best cutoff point is 0.54,the sensitivity is 92.50%,and the specificity is 91.25%.5.Referred to relevant literature,a questionnaire was designed to assess the elderly’s fall risk and investigate the current situation of falls among the elderly in Beijing communities.26.7%of the elderly had a history of falls,and 48.1%of the elderly were classified as high-risk groups.The main risk factors are advanced age,history of falls,use of walking aids,abnormal balance,sleep disorders,dizziness/vertigo,Visual impairment,hearing impairment,lower extremity arthritis,history of stroke,diabetes,depression,foot problems,arrhythmia,use of sedative and hypnotic drugs.Conclusion:1 The elderly have a higher incidence of falls,and there are many factors affecting falls2.The Logistic regression model constructed in this study is reasonable and scientific in its research process.The included indicators are comprehensive,simple and easy to measure,and the empirical model has good prediction efficiency,which can be used as a forecasting tool for the fall risk assessment of the elderly in China.3.The modeling process of this study can provide research ideas for the modeling of other chronic diseases.4.The application of this research model understands the fall risk level of the elderly in Beijing,and provides a scientific basis for the next step of disease control and health guidance.