Childhood obesity is considered a serious public health concern in the USA. In the last 30 years, childhood obesity has tripled, and recent data suggest that one in six American children is obese and one in three is overweight or obese( Reference Ogden, Carroll and Lawman 1 ). This dramatic increase in childhood overweight and obesity can be attributed to a number of factors, including individual child-level factors. Children’s dietary behaviours and their knowledge of nutrition-related information have been identified as primary mechanisms for weight gain. For instance, consumption of foods high in saturated fat and sugary beverages is related to paediatric weight gain( Reference Gillis, Kennedy and Gillis 2 ). In contrast, fruit and vegetable intake in childhood is associated with lower risk of being overweight and obese( Reference Lakkakula, Zanovec and Silverman 3 ). Unfortunately, most children in the USA do not meet current dietary recommendations of at least four servings of fruits and vegetables per day( 4 ), and in fact many children consume almost no servings of fruits or vegetables each day( Reference Siega-Riz, Deming and Reidy 5 ). Consumption of fruits and vegetables may be low in the USA because many children do not establish preferences for healthy foods early in life( Reference Birch 6 ). Food preferences develop in early childhood( Reference Birch 6 ) and preferences for fruits and vegetables have been positively linked with intake( Reference Birch 6 – Reference Domel and Thompson 8 ) and inversely associated with risk of obesity in children( Reference Domel, Thompson and Davis 7 ). Thus, targeting preferences in young children may be a promising approach for promoting consumption of fruits and vegetables and reducing risk for overweight and obesity. Although the home environment is a critical context for establishing young children’s food preferences and for promoting fruit and vegetable consumption, parents may unintentionally decrease children’s preferences for fruits and vegetables by decisions concerning food availability( Reference Benton 9 ). However, repeated exposure to healthy foods, even outside the home (e.g. at school), can lead to improvements in children’s preferences and willingness to consume these foods( Reference Benton 9 , Reference Anzman-Frasca, Savage and Marini 10 ).
Nutrition-related knowledge has also been identified as a mechanism for healthy weight in childhood, as this knowledge helps shape children’s dietary preferences and choices( Reference Contento, Balch and Bronner 11 , Reference Asakura, Todoriki and Sasaki 12 ). When children have an understanding of the benefits of including fruits and vegetables in their diets, they are more likely to choose and consume these foods( Reference Perez-Rodrigo and Aranceta 13 ). Schools can serve as a primary context for the development of healthy dietary behaviours and nutrition-related knowledge, and although evidence is mixed, previous research has found that these domains can be improved through school-based interventions( Reference Kandiah and Jones 14 – Reference Scherr, Linnell and Dharmar 18 ). For example, results from one study indicated that 2nd and 3rd graders who participated in a nutrition education programme delivered in elementary schools demonstrated greater improvements in dietary behaviours and nutrition knowledge than children in a control group( Reference Powers, Struempler and Guarino 17 ); however, in that study, nutrition educators, not classroom teachers, implemented the intervention. Other studies have found that programmes focused on school-based gardens may be effective in promoting dietary behaviours and knowledge( Reference Parmer, Salisbury-Glennon and Shannon 16 ), yet many schools do not have the space, resources or expertise to implement such programmes.
Despite the importance of developing healthy dietary behaviours and nutrition and health knowledge for children’s outcomes and the fact that some nutrition education programmes have been effective in elementary schools, few teachers incorporate nutrition education into their curricula( Reference Kann, Telljohann and Wooley 19 ). In a recent study, elementary-school teachers identified several reasons for not incorporating nutrition education into their classrooms, including a lack of suitable curriculum where lessons were feasible, easy to implement and short, and a lack of time to incorporate a stand-alone curriculum that did not align with required education standards( Reference Perera, Frei and Frei 20 ). Teachers in that study reported a preference for a curriculum that could be easily integrated into curricula from other subject areas (e.g. health, science) that included approximately ten sessions per year. Thus, the objective of the current study was to develop and evaluate a short (6-week) and feasible nutrition education curriculum that aligned with the National Health Education Standards designed to enhance children’s dietary behaviours and nutrition and health knowledge. Much of the previous research exploring the utility of school-based interventions, such as school gardens, has utilized non-experimental or quasi-experimental designs, limiting the validity and generalizability of study findings( Reference DeCosta, Møller and Bom-Frost 21 ); thus in the current study we employed a rigorous randomized controlled design to evaluate the curriculum. It was hypothesized that children who participated in the curriculum would exhibit greater improvements in a direct assessment of fruit and vegetable preferences and a nutrition and health knowledge survey relative to children in a control group.
Methods
Study design and procedures
In the current study we employed a cluster randomized controlled design. Using convenience sampling methods, three elementary schools in a Midwestern US state were recruited to participate in the study during the summer of 2016. In the autumn of 2016, teachers and children from all 2nd grade classrooms in the three schools were recruited to participate in the study. Teachers were recruited through an email with an informational letter describing study procedures. Teachers then provided written consent if they were willing to participate. Children were recruited to participate in the study through an informational letter sent to their parents describing study procedures. Parents then provided written consent if they were willing to allow their child to participate. Parents of 131 children returned consent forms and were included in the study sample. Children were, on average, 8 years old (53 % female) and were primarily Caucasian (90 %), with other races/ethnicities including 1 % Asian, 3 % African American, 2 % Hispanic, 1 % mixed race and 3 % unknown. Children’s parents had education levels ranging from some high school to a graduate degree, with an average level of education of a bachelor’s degree.
Following parent consent procedures, all participating children completed pre-test assessments, including a short survey on nutrition and health knowledge and a direct assessment of fruit and vegetable preferences administered by their teachers. Teachers were trained to administer the assessment in a standardized fashion and were told not to add any language beyond what was provided in the assessment script to avoid the introduction of any bias. After completion of pre-test assessments, all 2nd grade classrooms (n 10) were randomly assigned within school to be intervention or business-as-usual control classrooms (school 1: one intervention, one control classroom; school 2: one intervention, one control classroom; school 3: three intervention, three control classrooms; children: eighty-two in intervention classrooms, forty-nine in control classrooms). Children in intervention classrooms were not significantly different from children in control classrooms according to age, gender, parent education, ethnicity or pre-test assessments.
Teachers in intervention classrooms participated in a 2 h training session conducted by project staff either at their respective schools or at the research university where the project investigators were housed. Training consisted of a review of the curriculum day-by-day, including an outline of all activities. As a support, teachers received an intervention manual with all materials required to teach the programme, including class copies of all handouts. Teachers also received a $US 100 gift card to purchase fruits and vegetables necessary for food tastings. For consistency, questions regarding the programme or implementation were directed to project staff that conducted training.
Following training, teachers in intervention classrooms implemented a 6-week curriculum (two lessons per week) that was designed to promote children’s nutrition knowledge and fruit and vegetable preferences. Children in the control classrooms received no intervention and teachers carried on activities as usual. To establish fidelity of implementation, intervention teachers completed a brief evaluation at the end of each week’s instruction that included questions related to curriculum delivery. An example item included: ‘How many lessons did you teach this week?’ Items also included frequency questions, such as ‘How often did you deliver the lessons as was intended?’ These items were rated on a 4-point scale (1=‘rarely’, 4=‘always’). Following intervention implementation, all participating children completed post-test assessments that were identical to pre-test assessments. All study activities were reviewed and approved by the institutional review board at a research university through an expedited review process.
Intervention
The nutrition education curriculum used in the intervention was designed to improve children’s nutrition and health knowledge and preferences for fruits and vegetables. This intervention was grounded in Social Cognitive Theory( Reference Bandura 22 ), which emphasizes environmental inputs that contribute to children’s self-efficacy around dietary behaviours. As Bandura notes, from a social cognitive perspective, in order for school-based health promotion programmes to be effective, children must be equipped with the beliefs and skills needed to change their behaviours( Reference Bandura 23 ). Our curriculum provides children with age-appropriate background information on nutrition and health that is intended to change their belief systems around the benefits of eating healthy foods. Further, our curriculum gives children opportunities to develop the skills necessary to make healthy choices and try a variety of foods they may not have been exposed to before. Thus, our curriculum is in line with a Social Cognitive Theory approach.
The curriculum was developed during the summer of 2016 in collaboration with three 3rd grade teachers. Recruiting 3rd grade teachers was deliberate as we did not want to cut into our research pool of 2nd grade teachers, and 3rd grade teachers begin their teaching year with ‘ending’ 2nd grade students. Thus, they would understand the developmental stage of 2nd graders and be equipped to develop the curriculum. During a full-day workshop, the research team and teachers developed a 6-week curriculum, consisting of lessons implemented twice per week. The lessons were developed based on four existing programmes: MyPlate for grades 1–2 and 3–4, Two-Bite Club, and Put a Rainbow on Your Plate. As lessons were developed, they were aligned to the National Health Education Standards. This enabled teachers to cover topics required by the state while using the nutrition lessons. In addition to direct instruction on nutrition and health, a variety of age-appropriate activities were implemented throughout the programme including trivia, food tastings, projects and games.
Week 1 focused on the overarching questions, ‘What does it mean to be healthy?’ and ‘What does it mean to eat healthy?’ The MyPlate framework was highlighted. As a culminating activity for day 2, students labelled a blank copy of the MyPlate diagram after learning the components of healthy meals during both lessons of week 1. Week 2 emphasized the question, ‘Why is it important to eat a variety of foods from all food groups?’ The concept of nutrients was introduced and the book, Two-Bite Club, was read. The primary goal of the Two-Bite Club book is to encourage children to try at least ‘two bites’ of foods, and following this first reading children were challenged to try two bites of new foods as they were introduced throughout the rest of the nutrition curriculum. Week 3 focused on the questions, ‘What should I eat less of and why?’ and ‘What can I eat instead?’ This week was focused on teaching children strategies for replacing sweet or salty snacks with healthy choices. Week 3 also introduced the Put a Rainbow on Your Plate framework that was incorporated throughout the rest of the curriculum. Put a Rainbow on Your Plate encourages children to eat foods that are a variety of colours. In week 3, children learned the nutritional benefits of red foods while sampling a variety of red fruits and vegetables. Week 4 focused on the final question to be emphasized throughout the remainder of the programme: ‘Why should we eat fruits and vegetables?’ During this week, children learned the nutritional benefits of orange foods and sampled fruits and vegetables that are orange. The final project was also introduced to children this week. The final project served as an assessment of children’s understanding of key concepts introduced throughout the curriculum. The curriculum provided each class a choice from one of three projects: (i) children creating ‘Eat the Rainbow’ posters to be placed in the school cafeteria; (ii) children developing video commercials for various fruits and vegetables to be played on the school news; or (iii) children holding a farmers’ market (with pretend or real foods) to ‘sell’ to students in other grades or to school adults. Each project required children to report on the nutritional benefits of fruits and vegetables. Weeks 5 and 6 focused on the continuing theme of ‘Why should we eat fruits and vegetables?’ and emphasized the nutritional value and tasting of green and blue/purple foods, respectively. The final projects were presented in week 6.
Overall, intervention implementation fidelity was high. On average, over the course of 6 weeks, teachers reported delivering two lessons per week (mean 2·00 (sd 0·11)) that lasted approximately 15–20 min each, which was the targeted duration and dosage. Teachers also indicated that, on average, they implemented the lessons as intended (mean 3·65 (sd 0·42); 4-point scale). Finally, teachers reported that children were engaged in and enjoyed the lessons (mean 3·56 (sd 0·46) and mean 3·57 (sd 0·58), respectively; 4-point scale).
Measurement instruments
Parents completed a demographic survey including items related to children’s age, gender and ethnicity, as well as their own education level. Two child-level assessments were used in the present study to test the efficacy of the intervention. The first assessment was developed by the research team to align with nutrition and health content provided in the intervention. This survey consisted of fifteen questions. Sample items included: ‘What contains the most vitamin A per serving?’ (response options: 1=‘oranges’, 2=‘cantaloupe’) and ‘What you eat can make a difference in your chances of getting heart diseases or cancer’ (response options: 1=‘true’, 2=‘false’). Children received a score of 1 for each correct response and 0 for each incorrect response, with a possible range in scores from 0 to 15. The internal consistency (Cronbach’s α) for this outcome measure was 0·79.
The second assessment was a direct measure of children’s fruit and vegetable preferences that was modified from an assessment used in previous research( Reference Gillis, Kennedy and Gillis 2 ). Teachers administered this measure to participating children. During this assessment, teachers showed children forty-one coloured pictures of a variety of fruits and vegetables. While teachers were showing children a picture of a particular fruit or vegetable, they asked ‘What is your opinion of this [food]?’ Children were asked to respond with ‘I like it’ (score of 2), ‘I don’t like it’ (score of 1) or ‘I’ve never tried it’ (score of 0). Scores across all fruits and vegetables were summed, with a total range in scores from 0 to 82. The internal consistency for this outcome measure was 0·90.
Analytic procedures
Data analyses were conducted using the statistical software package Stata version 13.0. All outcome data were analysed for normality using skewness and kurtosis checks. Preliminary analyses, including t tests, were conducted to explore differences in outcome scores between the intervention and control children. Data were multilevel, with randomization occurring at the classroom level and child-level data being the unit of analysis. Intraclass correlations representing classroom variance in outcomes were 0·51 for the nutrition and health knowledge survey and 0·02 for the food preferences assessment. To account for this variance, multiple regression analyses that used the generalized Huber–White sandwich estimator to adjust se errors for non-independence (clustering by classroom) were conducted. Separate models were run for each outcome. Models included the following covariates: pre-test scores, age, gender, ethnicity and parent education. Statistical significance was determined using an α level of 0·05. There were very few missing data (<1 % for nutrition and health knowledge, <2 % for food preferences); however, full information maximum likelihood was employed to handle missing data.
Results
Descriptive statistics and correlations are presented in Tables 1 and 2. Intervention participation was marginally correlated with food preferences at post-test and significantly related to the nutrition and health survey. Skewness and kurtosis analyses indicated normal distributions for both outcomes. Results from analyses revealed that when controlling for the pre-test nutrition and health survey score, age, gender, ethnicity and parent education, children in the intervention condition demonstrated significantly higher scores on the nutrition and health survey compared with children in the control condition (β=0·47, P=0·001). Similarly, results suggested that when controlling for the pre-test fruit and vegetable preferences, age, gender, ethnicity and parent education, children who received the intervention showed greater preferences for fruits and vegetables at post-test than children who did not receive the intervention (β=0·19, P=0·003). See Table 3 for a summary of results and effect sizes.
Table 1 Descriptive characteristics of the full sample and intervention and control groups of elementary-school children participating in the nutrition education curriculum intervention in a Midwestern US state, autumn 2016
T1, pre-test assessment; Food pref, food preferences; T2, post-test assessment; Health, nutrition and health knowledge.
(*) P<0·10, ***P<0·001.
† t Tests comparing intervention and control groups.
‡ 1=8th grade or less, 2=some high school, 3=General Educational Development, 4=high school diploma, 5=some college, 6=associate’s degree, 7=bachelor’s degree, 8=master’s degree, 9=doctoral/postgraduate degree.
Table 2 Correlation matrix for all study variables (n 131)
T1, pre-test assessment; Food pref, food preferences; T2, post-test assessment; Health, nutrition and health knowledge.
(*) P<0·10, *P<0·05, **P<0·01, ***P<0·001.
† Child age measured in months.
‡ 1=8th grade or less, 2=some high school, 3=General Educational Development, 4=high school diploma, 5=some college, 6=associate’s degree, 7=bachelor’s degree, 8=master’s degree, 9=doctoral/postgraduate degree.
Table 3 Effects of the nutrition education curriculum intervention in a Midwestern US state on elementary-school children’s food preferences and nutrition and health knowledge, autumn 2016
*P<0·05, **P<0·01, ***P<0·001.
Discussion
The aim of the current study was to develop and evaluate a feasible nutrition education curriculum designed to enhance children’s dietary behaviours and nutrition and health knowledge. As expected, results from a randomized controlled trial indicated that 2nd grade children who participated in the curriculum demonstrated greater improvements in their fruit and vegetable preferences and nutrition and health knowledge relative to children in a control group. These findings suggest that a relatively short, 6-week intervention that aligns with educational standards has the ability to significantly enhance children’s outcomes and thus may be a more feasible option for teachers to incorporate into their classrooms than what is currently available.
Childhood obesity, and related health problems, has become a significant public health issue in the USA in recent years( Reference Ogden, Carroll and Lawman 1 ). In the current study, we chose to focus on designing an intervention that targeted children’s dietary behaviours (i.e. food preferences) and nutrition-related knowledge because of evidence suggesting that both have can a substantial impact on sustained behavioural change regarding consumption of healthy foods( Reference Domel, Thompson and Davis 7 , Reference Contento, Balch and Bronner 11 , Reference Asakura, Todoriki and Sasaki 12 ). For example, in a review, Blanchette and Brug noted that taste preferences are the among most critical determinants of children’s consumption of fruits and vegetables. Knowledge of recommended fruit and vegetable intakes was also identified as a key determinant of behaviour change( Reference Blanchette and Brug 24 ). As such, we developed a short and feasible nutrition education curriculum that aligns with the National Health Education Standards. Social Cognitive Theory was used as a basis for our intervention design; the intent was to equip children with the necessary skills and beliefs to effect actual behavioural changes( Reference Bandura 23 ). For example, by providing children with nutrition and health knowledge in a fun and engaging manner, the goal was to facilitate a sense of self-efficacy for making healthy choices. Our findings indicated that just two, 15–20 min lessons per week for 6 weeks were enjoyable for children and feasible for teachers. Further, participation in these lessons effected positive change in both targeted outcomes (food preferences and nutrition and health knowledge).
The findings with regard to nutrition education are in line with other school-based and nutrition-focused interventions( Reference Kandiah and Jones 14 – Reference Scherr, Linnell and Dharmar 18 ). For instance, in one study of similar duration, 5th grade children who participated in a nutrition education programme at school showed significantly greater gains in nutrition knowledge relative to a control group( Reference Kandiah and Jones 14 ). Similarly, in another study, 2nd and 3rd grade children exposed to six weekly nutrition education lessons demonstrated greater improvements in nutrition knowledge compared with children in a control condition( Reference Powers, Struempler and Guarino 17 ). There have been few school-based and nutrition-focused interventions that have been effective for promoting children’s fruit and vegetable preferences. For example, in a multicomponent, school-based intervention study, although there were intervention effects on nutrition knowledge and vegetable identification, there were no significant effects on children’s food preferences( Reference Scherr, Linnell and Dharmar 18 ). The majority of school-based interventions that have been effective in improving food preferences include a garden component. Morris and Zidenberg-Cherr found that although a nutrition education curriculum did produce favourable effects on food preferences for 4th grade children, effects were greater and sustained when garden activities were included( Reference Morris and Zidenberg-Cherr 25 ). Although garden-based interventions have promise for improving food preferences, they may not be feasible for many schools. Gardens can be expensive to build and time-consuming to maintain, and some schools may not have the resources or space for this endeavour. Thus, one goal of the current study was to include other activities that may promote nutrition knowledge and food preferences that did not include a garden component.
One key ingredient of our curriculum that may have facilitated positive effects on both dietary behaviours and nutrition and health knowledge is the inclusion of tastings. For young children, nutrition education may not always be enough to effect actual change in children’s dietary behaviours; however, exposure to healthy fruits and vegetables with tastings has been shown to promote preferences for these foods( Reference Wardle, Herrera and Cooke 26 ). As part of our nutrition education curriculum, children were exposed to and tasted a variety of fruits and vegetables that were introduced in fun and engaging ways. For example, in week 2, children were introduced to the Two-Bite Club via a storybook reading that encourages children to try at least two bites of new foods, just like the characters in the Two-Bite Club book. Then, children were given new foods to taste (e.g. pomegranate, butternut squash, okra). This inclusion of tastings along with nutrition education may be critical for promoting healthy dietary behaviours in young children. It will be important for future research to test this hypothesis.
An additional strength of the present study is the potential for the curriculum to be integrated into other content domains. Although the curriculum primarily targets nutritional components of the National Health Education Standards, it could also meet some of the National Health Science Standards. For example, one of the foci of the ‘Science in Personal and Social Perspectives’ standard is personal health. Teachers could use the curriculum to meet this particular standard in addition to meeting health education standards.
Although the current study has a number of strengths (e.g. randomized controlled design), limitations must be noted. First, our sample size was relatively small and homogeneous. It will be critical for future research to replicate our findings with larger and more diverse samples. Second, our targeted outcomes were limited to direct assessment of food preferences and child report of nutrition and health knowledge. Although both of these outcomes have been associated with consumption in previous research( Reference Ogden, Carroll and Lawman 1 , Reference Contento, Balch and Bronner 11 , Reference Asakura, Todoriki and Sasaki 12 ), future research evaluating the curriculum would benefit from the inclusion of pre- and post-test assessments of fruit and vegetable consumption. Finally, although the short-term intervention was found to effective, we did not have a long-term follow-up to assess the extent to which the effects were maintained. It is possible that the intensity of the intervention was not enough to lead to sustained behaviour change. Future research should test whether a more holistic approach to curriculum that more fully integrates nutrition into all areas of education (e.g. health, maths, science, arts) would lead to even stronger effects in the short and long term.
Despite these limitations, our study supports the inclusion and effectiveness of a nutrition education programme for promoting young children’s fruit and vegetable preferences and nutrition and health knowledge. Preferences for and knowledge of fruits and vegetables predict children’s consumption of healthy foods( Reference Ogden, Carroll and Lawman 1 , Reference Contento, Balch and Bronner 11 ). Thus, exposing children to healthy foods and providing an education on the value of these foods early in life may help prevent the onset of overweight and obesity, laying the foundation for lifelong health and well-being.
Acknowledgements
Financial support: This study was supported by a grant from the Indiana State Department of Agriculture, Specialty Crop Block Grant Program (EDS #A337-15-SCBG-14-006). The funder had no role in the design, analysis or writing of this article. Conflict of interest: None. Authorship: S.A.S. designed the study, recruited participants, collected data, trained teachers, analysed the data and led manuscript preparation. L.M.B. managed data, analysed the data and assisted in manuscript preparation. I.K. managed data, analysed the data and assisted in manuscript preparation. L.K. recruited participants, collected data, trained teachers and assisted in manuscript preparation. B.K. helped design the study, collected data and assisted in manuscript preparation. T.B. managed the grant and assisted in manuscript preparation. Ethics of human subject participation: This study was conducted according to the guidelines laid down in the Declaration of Helsinki and all procedures involving human subjects were approved by the Purdue University Institutional Review Board. Written (from parents) and verbal (from children) informed consent was obtained from all subjects. Verbal consent was witnessed and formally recorded.
