Development Of Semi-quantitative Food Frequency Questionnaire And Dietary Patterns Research For Preschool Children In Northwest China

ObjectiveA reasonable diet is the foundation of children's growth and physical health.The preschool period is a critical period for children's growth and development.The nutritional status of this period directly determines their growth and development level and future physical health.The unique geographical location and climatic characteristics of the Northwest region determine its unique food culture and dietary pattern.However,due to the lack of high reliability and validity of dietary survey and evaluation tools for preschool children in Northwest China,the dietary intake,nutritional status,and dietary patterns of preschool children in Northwestern China are lacking,as well as the relationship between dietary patterns and health.The lack of a comprehensive and objective understanding of the relationship has affected the scientific formulation of interventions related to the physical health of preschool children.Base on this,our research has the following three purposes:Firstly,using a standardized questionnaire compilation process to develop a semi-quantitative food frequency questionnaire for preschool children in Northwest China with good reliability and validity;Secondly,Using multiple multivariate statistical methods to extract the dietary patterns of preschool children in Northwest China.Lastly,multi-factor analysis was used to explore the dietary patterns relationship with health index,and to study the effects of dietary patterns on health indicators quantitatively,such as obesity,sleep and other indicators.Methods1.Construction of semi-quantitative food frequency questionnaire frameworkBased on relevant retrospective learning,the concept of food frequency for preschool children is preliminarily determined,which including the different content,form,dietary characteristics,and influencing factors of food frequency for preschool children,also taking into account the diet culture and habits of the northwestern region,to build a framework of diet frequency measurement for preschool children.2.Development of initial food frequency questionnaireAccording to the framework,a food list for preschool children is established through a 3-day 24-hour diet review method and a qualitative interview with the target group.An initial questionnaire is formed through expert consultation,focus group discussions,surveys of the main caregivers and target groups.3.Development of trial food frequency questionnaireAccording to the convenience sampling method,a sample of 218 children was selected from 3 districts of Xi'an,Lanzhou and Yinchuan.Use the initial questionnaire to collect data on dietary intake types and intakes of preschool children.On this basis,with reference to the"Chinese Food Composition Table(2017 Edition)"and related literature,19 nutrients were determined as basic nutrients under the guidance of relevant experts.The cumulative percentile method used 90%of the basic nutrients as the screening standard,and the multiple regression analysis was based on the determination coefficient R~2accumulation of 0.9 as the basis selection criteria.Comprehensive consideration of the expert opinion to revise the items to form a trial questionnaire.4.Development of formal food frequency questionnaireBy means of convenient sampling,266 children aged from 3 to 6 years were selected from 5 regions,including Xi'an,Lanzhou,Yinchuan,Xining and Urumqi.A trial questionnaire was used to investigate.The classical test theory including factor analysis method,Cronbach's coefficient method and discrete trend method are used to comprehensively analyze each item and eliminate the items which do not meet the conditions.Exploratory factor analysis was used to determine the dimensions of the questionnaire,and the items were revised according to the expert opinions.Finally,a formal questionnaire was formed--"Preschooler's Semi-quantitative Food Frequency Questionnaire in northwest China"(PSFFQ).5.Evaluation of the formal questionnaire326 preschool children were selected from Xi'an,Lanzhou and Yinchuan for food frequency investigation,Using the PSFFQ and"3d24h Dietary Review Questionnaire for Preschool Children"as questionnaire tools to conduct food frequency survey.The PSFFQ questionnaire was surveyed twice with an interval of 4 weeks.Between the two PSFFQ diet surveys,a 3d24h diet review survey was conducted.The retest reliability of the energy and nutrient analysis questionnaire was obtained from two PSFFQ surveys,and the energy and nutrient calculation obtained from the second PSFFQ survey data to divide the half-reliability and Cronbach's a coefficient.Using the energy and nutrient data obtained from the 3d24h diet review as the standard,the average of PSFFQ diet data was used to verify the validity of the calibration.Using the second PSFFQ diet survey food intake frequency data for structural validity test,and the content validity ratio(CVR)to check the content validity of the questionnaire.6.Dietary pattern extraction of preschoolers in northwest ChinaPSFFQ was used to collect 541 dietary intake data of children who aged 3?6 in Xi'an,Lanzhou,Xining,Yinchuan and Urumq with convenience sampling method.We used the latent profile analysis,cluster analysis and factor analysis combined with cluster analysis to extract the preschool children's dietary patterns respectively.7.Analysis of influencing factors of preschool children's diet patternUsing disorder multi-category logistic regression method,different dietary patterns as dependent variables,demographic characteristics of preschool children and factors such as the number of meals outside,outdoor exercise time,screen time and other factors are used as independent variables to analyze the influence of preschool children's dietary patterns factor.8.Relationship between dietary patterns and health statusDifferent diet patterns which extracted by factor analysis combined with cluster analysis were used as independent variables,and whether preschool children are obese,whether their sleep quality is normal,whether they have respiratory diseases in the past month,whether they have indigestion in the past month were used as dependent variables.Using binary logistic regression to analyze the correlation between the demographic characteristics and dietary patterns of preschool children and various health indicators.Results1.A preliminary draft of 166 types of questionnaire covering 11 major categories and 19 basic nutrients was established.After preliminary testing by kindergarten teachers and nutrition experts,preliminary screening and evaluation of food list items were conducted to form an initial questionnaire which have 153 items in 11 dimensions.2.According to the results of the first round of preliminary semi-quantitative food frequency questionnaire of 218 preschool children,the items of the questionnaire were screened again by the accumulative percentile method and multiple regression analysis,and a trial questionnaire containing 87 kinds of food was formed.According to the second round of 266 preschoolers'trial questionnaire survey data,the items were screened again,and factor analysis was used to extract 9 common factors,combined with experts'suggestions,and finally a formal questionnaire with 66 items in 9 categories was formed.They are respectively named as staple food,meat and its products,eggs and milk and its products,beans and its products,fruits,dark vegetables,light vegetables,mushroom vegetables and beverages,the cumulative variance contribution rate of 9common factors was 53.7%.3.The results of measurement attribute evaluation of formal questionnaire show that the questionnaire has good reliability and validity.Cronbach's a coefficient was 0.819,Cronbach's a coefficient of energy and each nutrient is between 0.780?0.829.The half-reliability coefficient was 0.869 and the half-half reliability coefficient of energy and each nutrient is between 0.723 and 0.942.The reliability correlation coefficient of retest was 0.542,and the reliability correlation coefficient of each nutrient was between 0.282and 0.882.In the validity test the contribution rate of exploratory factor analysis's 9 factor variance reached 64.42%,the factor loading of confirmatory factor analysis was basically greater than 0.4,and the coefficient of calibration validity analysis was 0.368?0.814,with an average of 0.613.The content validity ratio CVR=0.8.4.In order to avoid the loss of information for food with low consumption rate when extracted the dietary pattern,9 categories of 66 foods were classified into 20 categories according to food types,nutrient content,to extract dietary patterns.Dietary patterns were extracted from dietary survey data of preschool children by latent profile analysis.According to the minimum values of AIC,BIC and a BIC,the LMR and BLRT tests were statistically significant(P<0.05),the 3-category model was selected as the ideal model.In accordance with the potential category probability of 3 potential categories in 20categories food,the dietary patterns of preschoolers were named as"high quality dietary patterns,regular dietary patterns and low quality dietary patterns".Their distribution among preschoolers was 9.22%,57.38%and 33.41%,respectively.5.K-means cluster analysis method for extracting dietary patterns of preschool children's food intake.Under the guidance of nutrition experts,the number of clusters is preset to 3,and the maximum number of categories is set to 20 for rapid cluster analysis.They are named according to the types of food contained in different types of dietary patterns,which are traditional,healthy,and high-sugar three types of dietary patterns.Its distribution among preschool children is 48.80%,33.09%and 18.11%,respectively.6.An exploratory factor analysis of the average daily intake of 20 kinds of food for preschool children.On the basis of referring to the scree plot and characteristic values,four common factors were reserved according to the suggestions of nutrition experts,and the contribution rate of the cumulative variance was 62,17%.On this basis,K-means clustering analysis was used for cluster analysis.According to the results of K-means cluster analysis,the four dietary pattern were named staple food,high protein,healthy and high sugar dietary patterns.The distribution among preschool children is 41.77%,26.99%,23.66%and 7.58%.The staple food-based dietary patterns consume more foods such as cereals and fillings,which is traditional dietary pattern in northwest China.However,this mode intake less fat,dietary fiber,vitamin C,and vitamin E(P<0.05).High protein diet pattern consumed more milk,yogurt and eggs,which were negatively correlated with dietary fiber,vitamin A,vitamin C and other nutrients(P<0.05).The healthy dietary patterns include more food such as meat,livestock,poultry,aquatic products,animal organs,and dark-colored vegetables,which are negatively correlated with carbohydrates(P<0.05).The high-sugar diet model consumes more foods such as sweets and beverages,which are negatively correlated with proteins,dietary fiber,vitamin E,and calcium(P<0.05).7.Using multivariate disorderly multi-class logistic regression to analyze the influencing factors of different dietary patterns,children aged 6?7 were more likely to choose high-sugar dietary patterns(OR=3.263,95%CI:1.834-11.741,P<0.05).Mothers with a college education level or above are more likely to choose a high-protein diet(OR=2.263,95%CI:1.245?8.456,P<0.05),and a staple food type with an average monthly household income of more than 8,000 yuan(OR=0.642,95%CI):0.165?0.914,P<0.05)and the protective factors of high-sugar diet(OR=0.620,95%CI:0.138?0.798,P<0.05);the family structure is based on a family of three,It's easier to choose a staple food diet when live with grandparents(OR=2.783,95%CI:1.392?6.735,P<0.05);those who have eaten out more than 8 times in the past month are more likely to choose a high-sugar diet(OR=2.248,95%CI:1.806?10.570,P<0.05),it is easier to choose a high-sugar diet if the screen time is more than 2 hours per day(OR=3.217,95%CI:1.579?9.732,P<0.05)8.The dietary patterns that affect overweight and obesity in preschool children are the staple food diet(OR=1.762,95%CI:1.212?2.561,P<0.05)and the high-sugar diet(OR=1.690,95%CI:1.021?2.810,P<0.05),children with high-sugar diet have poor sleep quality(OR=2.357,95%CI:1.014?4.078,P<0.05),children with high-protein diet may have better sleep quality(OR=0.753,95%CI:0.109?0.863,P<0.05),and high-protein diet are not susceptible to respiratory infections(OR=0.465,95%CI:0.251?0.893,P<0.05).Conclusions1.The semi-quantitative food frequency questionnaire for preschool children prepared in this study included 9 categories of 66 kinds of food,including stable food,meat and its products,eggs,milk and its products,beans and its products,fruits,dark vegetables,light-colored vegetables,mushroom and beverages.This questionnaire has good quality and can be used for the nutrition status survey of preschool children in northwest China.2.Factor score,a continuous variable that can be analyzed repeatedly in factor analysis,is combined with the advantages of clustering analysis to classify individuals into different types of dietary patterns.The research method for extracting dietary patterns for preschool children has a certain scientific nature.When latent profile analysis was used to extract children's dietary patterns,there were such problems as high data requirements,group characteristics could not represent individual categories,and uncertainty of model estimation standards in nutrition.Therefore,the dietary patterns extracted by this method in this study need to be further verified.3.Combining factor analysis and cluster analysis,the dietary patterns of preschool children can be divided into four types:Staple food,high protein,healthy and high sugar dietary patterns.It was found that the intake of nutrients such as dietary fiber,vitamin E,and vitamin C in the staple food and high-protein diet was insufficient,the carbohydrate intake in the healthy diet was less,and the fat,dietary fiber,calcium and other nutrients in the high-sugar diet were inadequate.Therefore,children's dietary guidance and daily feeding should be based on different types of dietary patterns to increase the relevant nutrient intake in order to achieve the purpose of reasonable nutrition and balanced diet.4.Through research on the factors affecting the dietary patterns of preschool children,it is found that older children are more likely to choose high-sugar diets,children with higher education levels of mothers are more likely to choose high-protein diets,and preschool children from high-income families are easier to choose high-protein dietary mode.Preschool children living with parents and grandparents choose staple food dietary mode.Children with more outside meals are more likely to choose high-sugar dietary mode.The longer they spend watching TV,mobile phones and other screens,the children are more inclined to choose a high-sugar diet.5.Through the study of the correlation between dietary patterns and health indicators of preschool children,preschoolers who tended to follow staple food dietary patterns and high-sugar dietary pattern were more likely to be overweight and obese.Children who prefer high-protein diets have better sleep quality,and children who prefer high-sugar diets have poor sleep quality.Children who prefer a high-protein diet are less susceptible to respiratory infections than other children.Innovations1.For the first time in China,preschool children in northwest China were taken as research objects,and according to the standard questionnaire development process,a tool for the investigation of dietary status with reliability and validity,was developed,“Semi-quantitative food frequency questionnaire for preschool children”.2.It is the first to use factor analysis combine cluster analysis to explore four main dietary patterns for preschool children in Northwest China,namely,healthy diet,staple food,high protein diet and high sugar diet.3.The study found that high-sugar and staple food dietary pattern were a key factor in overweight/obesity and sleep quality in preschool children.The high-protein dietary patterns are beneficial factors in promoting sleep quality and reducing the incidence of respiratory tract infections.Therefore,preschool children should reduce their intake of unhealthy foods represented by beverages,and increase the promotion of high-protein dietary patterns.