Construction Of Guangxi Healthy And Longevity Diet Model Recommendation System And In Vitro Fermentation Verification Of Its Operation Effect

Diet is the foundation of life and the guarantee of health and longevity.A large number of studies have found that diet will have a fundamental impact on health and longevity.Guangxi has a remarkable phenomenon of longevity.It is of great significance to construct a nutritional diet data platform with Guangxi characteristics from the perspective of "drawing lessons from longevity" and apply it to dietary guidance of ordinary middle-aged and elderly people.This research team with the investigation and the testing data of guangxi longevity and the general population(diet characteristics,physical examination index,intestinal flora,metabolites)as the foundation,first the random forest(RF)algorithm was used to construct a guangxi longevity idealized health state model,then further refine diet characteristics and the quantitative relationship between health indicators,based on this,guangxi longevity diet recipe and personalized diet recommendation method are designed and implemented in Guangxi longevity nutrition diet data platform.Further,in vitro fermentation method was used to verify the probiotic properties of the recommended personalized diet.Through the above research,the aim is to provide reference and help for the improvement of the health status of the general population through the guidance of personalized diet program.The main results are as follows:1.Based on random forest classification algorithm(RF),an ideal health state model of guangxi longevity population was constructed.Three RF classification models were established based on the physical examination index,intestinal microflora,metabolites and other data collected from the long-lived population and the general population through data pretreatment,and the area under the subject characteristic curve(AUC)value was used to evaluate the model.The results show that the average AUC values of ten cross-validation times are all greater than0.95,indicating that the models have good generalization ability.A total of 24 key factors of longevity in Guangxi were screened out and the optimal trend values of each key factor were analyzed by kernel density curve.In order to provide more detailed target reference value for the improvement of the health status of the general population,an ideal health status model of guangxi long-lived population was established based on the statistical range and optimal trend value of the key factors.2.Random forest regression algorithm was used to analyze the quantitative relationship between dietary characteristics and health indicators in long-lived population.Further refine quantitative relationships between dietary characteristics and blood key factors in idealized health status models.The comprehensive health index(HCI)of the long-lived population in Guangxi was constructed by using health index classification and scoring method and entropy weight method.Subsequently,RF regression modeling was conducted for the relationship between dietary characteristics of longevity population and HCI.Accuracy was used to evaluate the results of the model.Through the quantitative relationship among the indicators illustrated by the model,it was found that the RF regression model between dietary characteristics and health indicators of all age groups had high accuracy.Among them,the regression accuracy of food and nutrient intake to HCI of centenarian group was the highest(93.37%)and 93.40%,respectively.According to the intensity of the impact of dietary characteristics on health,it was found that high intake of fish and shrimp,low intake of red meat,high intake of vitamin A and high intake of dietary fiber were the unique dietary characteristics of the centenarian group and 90 s group,which had A positive impact on their HCI,which provided A reference for improving human health through diet.Compared with the domestic and foreign studies on the correlation between diet and health,this chapter explains the quantification of the relationship between diet characteristics and health indicators from a new perspective.3.Preliminary construction of diet-based data platform of Guangxi health and longevity recommendation program.Based on the research data and information on the comprehensive impact of diet,vitamin,dietary fiber and meat intake characteristics on health of longevity population,this paper designed the longevity nutrition diet recipe and personalized diet recommendation method of Guangxi,using Spring Boot,MyBatis and other professional technologies.The data platform of Guangxi longevity nutrition diet was constructed and displayed in the form of Web page.At the user level,functional requirements such as user login and personalized diet plan recommendation are realized.At the administrator level,functional requirements such as user management,supplement and improvement of standard diet plan are realized.The core functions of users and administrators have passed the black box test.The construction of guangxi longevity nutrition diet data platform provides a new means of support for dietary nutrition guidance to improve the health status of ordinary middle-aged and elderly people in Guangxi.4.The platform runs in vitro fermentation probiotic characteristics test and feedback of recommended personalized diet plan.In vitro fermentation experiment was used to test the concentrations of 6short-chain fatty acids and the relative expression levels of 4 intestinal bacteria in fermentation broth at different sampling points(0,6,12 and24h)after food pretreatment,gastrointestinal digestion and colon fermentation in vitro.The results showed that at 24 h of fermentation,compared with the normal diet group(B)and the normal diet group(C),the personalized diet group(A)significantly promoted the production of acetic acid,propionic acid,isobutyric acid,butyric acid and isopentanoic acid(P(27)0.05),and significantly promoted the growth of Bifidobacteria and Lactobacillus(P(27)0.05).The inhibition degree of Escherichia coli and Bacteroidetes was the highest.The results show that the personalized diet program recommended by the data platform has good probiotics compared with the ordinary diet program.