Development And Application Of Nutrition Literacy Questionnaire For Middle School Students In Chongqing

Objective: This study aims to develop a valid and reliable questionnaire to assess nutrition literacy in middle school students in Chongqing and its influencing factors,to provide a research basis for the guidance of nutrition education for middle school students.Methods: This study consisted of two parts.Part I: Developing and validation of the nutrition literacy questionnaire.Establishing a theoretical framework and initial Chongqing middle school student nutrition literacy scale(CM-NLS)item pool based on the literature review and group discussion.Modifying the CML-NLS with two rounds of Delphi expert consultation and pre-test.Verifying its validity with Classical Test Theory(CTT)and exploratory factor analysis(EFA)after a pilot study.SPSS 20.0 and STATA 16.0 software were used for confirmatory factor analysis(CFA)and verify its reliability,Amos 23.0software was used to evaluate the structural validity.Part II: Application of the questionnaire.An anonymous online survey study,convenience sampling,was conducted on middle school students(grade 7,8,10,11)in 29 districts of Chongqing in September 2020.The self-constructed questionnaire was used to probe three parts including informed consent,demographic information,the channels and barriers for students to obtain nutrition information and CML-NLS which containing3 domains(functional,interactive,critical),6 skills(obtain,understand,application,interactive,media literacy,critical),and 52 items,with a total score of 167 points.The final score was converted into a centesimal measure for the total score of nutritional literacy,three domains and six skills.The higher was the better nutrition literacy of middle school students.Divide the total score(0-100)into high and low two groups by median.STATA 16.0 software was used for data analysis.The χ2 test was used to compare the nutritional literacy of middle school students with different demographic characteristics.Binary logistic regression was used to analyze the influencing factors of nutritional literacy.Results: Part I: A total of 18 experts participated in the Delphi method.The expert consultations’ response rates were above 0.7,the coordination coefficient was 0.612(P<0.001)and 0.431(P<0.001)in the first and second round,respectively.The evaluation indexes system comprised 3principal indicators(domains),6 sub-indicators(skills)and 59 items.Finally,52 items were retained based on the CTT,EFA,CFA and professional judgment of the CML-NLS.Confirmatory factor analysis showed that χ2,RMSEA,GFI and AGFI were 0.000,0.063,0.822,0.790,respectively,which indicated that the model fitted indexes reached the standard.The total Cronbach’s α of the scale was 0.849,and the Cronbach’sα coefficients of the domains of functional,interactive,or critical literacy were 0.826,0.942 and 0.938,respectively,which indicated that CM-NLS with reasonable reliability and validity.Part II: A total of 20372 valid questionnaires were collected.The average age of the students was 14.39±1.90 years old.The median CMLNLS score of students was 61.08.The scores of the domains from high to low were interactive(70.00),functional(67.68)and critical literacy(45.83),respectively.The six skills scored from high to low were application(72.73),interactive(70.00),obtain(66.67),understand(62.50),critical(50.00),and media literacy(40.63).The analysis of nutrition literacy’ influencing factors showed that when controlling for other confounding factors,minority students were less likely to have a high level of nutritional literacy than Han students(OR=0.687,95%CI=0.622～0.759),high school students(grade 10 and 11)were less likely to have a high level of nutritional literacy than those who were grade7(OR=0.495,95%CI=0.452～0.543;OR=0.511,95%CI=0.467～0.559).Those students who were living in rural had less likely to have a high level of nutritional literacy than those in town(OR=0.850,95%CI=0.791～0.913).The students whose family caregiver model was only grandparents/other caregiver had lower scores than those of grand-parent mixed caregiver models(OR=0.783,95%CI=0.684～0.895;OR=0.831,95%CI=0.729～0.948).The scores of students with higher educational level fathers(junior middle school/ senior high)were higher than those with father’ educational level was elementary school or below(OR=1.119,95%CI=1.032～1.214;OR=1.190,95%CI=1.063～1.333).The scores of students with higher educational level mothers(junior middle school/ senior high/ college degree and above)were higher than those with mother’ educational level was elementary school or below(OR=1.219,95%CI=1.129～1.316;OR=1.243,95%CI=1.107～1.395;OR=1.296,95%CI=1.094～1.534).Overweight/obese were lower than those with normal BMI(OR=0.935,95%CI=0.879～0.995).Conclusion: The Chongqing middle school student nutrition literacy scale(CM-NLS)was a valid and reliable instrument.At present,the nutritional literacy level of middle school students in Chongqing was at the middle level,and their critical nutrition literacy needed to be improved.These results highlighted groups of ethnic minorities,senior high school,rural areas,parents with low educational level,mixed parenting model of non-grandparents-parents,overweight and obese in the school population who were at higher risk of having lower nutrition literacy level.Therefore,the government,schools,families,media communicators,and food companies,as stakeholders in student nutrition education,should all play an active role to understand nutrition education,improve systemic nutrition education courses,strengthen family education,and actively promote and guide correct nutrition knowledge.Improve students’ nutritional literacy.