Construction Of A Discriminant Model For Simple Indicators Of Fall Risk Among The Elderly Aged 60-70

Objective: this study takes the relevant simple test methods of fall risk assessment for the elderly as a starting point,and establishes a practical and easy to popularize model by studying the relationship between the physical function test results of the elderly and the occurrence of falls.It is convenient for large-scale fall risk screening model for the elderly to find the elderly with high fall risk as early as possible,and take intervention measures to reduce the occurrence of falls and improve the quality of life of the elderly.To reduce the medical and health expenditure caused by falls among the elderly in our country.Methods: 150 healthy people aged 60-70 years old in Tianjin were divided into fall group and non-fall group according to the fall history survey.The visual function of the elderly was evaluated by visual acuity and visual sensitivity,the number of sitting in 30 seconds and the average peak moment of the knee extensor and flexor of the knee joint were evaluated to evaluate the lower limb muscle strength,walking speed,stride frequency,stride length and stride width of the elderly.the gait characteristics of the elderly,eye-closed standing on one foot,functional extension test and Berg scale were used to evaluate the balance ability of the elderly,and the response ability of the elderly was evaluated by falling scale response time.The statistical method is used to cluster the test indexes,and the typical evaluation indexes are selected in each category.According to the discriminant rules,a discriminant function model is constructed,and the prediction effect of the model is verified.Results: 1.By clustering the test indicators,the analysis results showed that they could be divided into four categories: balance function,lower limb muscle strength,visual-mobility ability,and reaction ability.2.The movement principal component analysis method carries on the factor analysis,and evaluates the evaluation index according to the eigenvalue,variance contribution rate and cumulative contribution rate of the principal component.the final typical indicators are: standing on one foot with eyes closed,step length,step frequency,10% gray visual sensitivity,timing standing up and walking test,step width,30 s sitting up test,falling ruler reaction time.3.The results of independent sample T test showed that the gray visual sensitivity of 10% in the fall group was significantly lower than that in the non-fall group(P < 0.05),and the time of standing up and walking in the fall group was significantly longer than that in the non-fall group(P < 0.05).The time of eye-closing and standing on one foot in the fall group was significantly lower than that in the non-fall group(P < 0.05),and the number of sitting in 30 seconds in the fall group was significantly less than that in the non-fall group(P < 0.05).There was no significant difference in step frequency between fall group and non-fall group(P >0.05),there was no significant difference in step length between fall group and non-fall group(P > 0.05),and there was no significant difference in step width between fall group and non-fall group(P > 0.05).There was no significant difference in fall reaction time between fall group and non-fall group(P > 0.05).4.Fisher discriminant function and Bayes discriminant function were constructed respectively.According to the standardized function coefficient,the order of index weights from large to small was obtained as follows: standing time on one leg with eyes closed,timed standing walking test,gender,10% grayscale visual sensitivity,30 s number of sitting.5.The verification results of the simple index discriminant model show that the total prediction accuracy of the internal verification of the model is 84.7%,the total prediction accuracy of cross-validation is 84.0%.External verification recruits 50 elderly people who meet the experimental conditions and tests them.The results show that the correct rate is 86%,and the verification result of the model is good.Conclusion: 1.The construction of a simple index discrimination model for the elderly has been completed,and it has been proved that the model has good predictability.2.By including the indicators of the discriminant function model,it can be found that 10% gray visual sensitivity,standing on one foot with eyes closed,timing stand-up test,30 s sit-up test and cross-step length are closely related to falls in the elderly.at the same time,it is also suggested that we should intervene in the balance function,visual function and lower limb muscle strength of the elderly to improve the functional level of the elderly and reduce the incidence of falls.