Study selection

Our literature search yielded 7809 publications between 1997 and 2020 (PubMed = 1057, SCOPUS = 1880, Web of science = 4535, Cochrane = 98, Pro Quest = 123, and Google Scholar = 116). After removing duplicates, 102 articles were screened based on title and abstract review. Of these, 64 publications were excluded for the following reasons: no full text available (n 29), thesis (n 26), the paper was not in English (n 2), book, conference abstract (n 5) and review (n 2). The full texts of the remaining thirty-eight publications were retrieved for further assessment, of which nineteen failed to meet the inclusion criteria. The main reason for excluding full texts was that they were not school-based interventions (Fig. 1, online Supplementary Table S2). Finally, nineteen articles were included, such that their characteristics are summarised in Table 2. The quality assessment of each of these studies is depicted in Fig. 2.

Fig. 1. PRISMA diagram.

Table 2. Key characteristics of reviewed studies (n 19)

FL, food literacy; NL, nutrition literacy; NS, not stated; SCT, Social Cognitive Theory; TPB, Theory of Planned Behavior.

Fig. 2. Quality assessment (using the EPHPP) of reviewed studies (n 19).

Study characteristics

The main theoretical models of behaviour change used in developing food/nutrition literacy interventions were Social Cognitive Theory (n 5, 26·31 %)^{(Reference Miller, Franzen-Castle and Aguirre31–Reference Scherr, Linnell and Dharmar35)} and Theory of Planned Behavior (n 1, 5·26 %)^{(Reference Beckman and Smith36)}. Theory-based interventions mainly improved functional food/nutrition literacy (Table 3).

Table 3. Summary of intervention description in terms of content, facilitators, cooking course association setting and its effect on the FNLIT dimensions and its components by the quality level of study

Four studies (21 %) were randomised controlled trials^{(Reference Scherr, Linnell and Dharmar35,Reference Block, Gibbs and Staiger37–Reference Townsend, Johns and Shilts39)} , and four (21 %) used a case–control design^{(Reference Beckman and Smith36,Reference McAleese and Rankin40–Reference Katz, Katz and Treu42)} . In three studies (15·78 %), two groups were compared pre- and post-test^{(Reference Morgan, Warren and Lubans33,Reference Treu, Doughty and Reynolds43,Reference Barnick44)} , but most studies (n 8, 42·10 %)^{(Reference Miller, Franzen-Castle and Aguirre31,Reference Gavaravarapu, Saha and Vemula32,Reference Wolfe and Dollahite34,Reference Thonney and Bisogni45–Reference Cunningham-Sabo and Lohse49)} used the same group tested pre- and post-intervention.

Fifteen out of nineteen studies (78·94 %) had not used a valid scale to measure FNLIT and its components. Only four studies (%) measured food label literacy by valid measures^{(Reference Gavaravarapu, Saha and Vemula32,Reference Treu, Doughty and Reynolds43,Reference Hawthorne, Moreland and Griffin48,Reference Katz, Treu and Ayettey50)} . In one study, a change in knowledge of food labelling was assessed by asking individuals whether a food label was present on a product^{(Reference Gavaravarapu, Saha and Vemula32)}. Validated multi-item ‘food label literacy’ tools to evaluate the food label literacy of students were used only by two studies^{(Reference Treu, Doughty and Reynolds43,Reference Katz, Treu and Ayettey50)} . Treu et al. ^{(Reference Treu, Doughty and Reynolds43)} evaluated knowledge of healthy food choices in the form of food label literacy in school-aged children by the Food Label Literacy and Nutrition Knowledge (FLLANK) questionnaire, which previously underwent validation testing in the Independence School District (ISD)^{(Reference Reynolds, Treu and Njike51)}.

Eighteen of the nineteen studies were set in high-income countries, as classified by the World Bank economic classification^{(Reference Fantom and Serajuddin52)}. Of these, fifteen studies were conducted in the USA^{(Reference Miller, Franzen-Castle and Aguirre31,Reference Morgan, Warren and Lubans33–Reference Beckman and Smith36,Reference Gold, Larson and Tucker38–Reference McAleese and Rankin40,Reference Katz, Katz and Treu42–46,Reference Hawthorne, Moreland and Griffin48–Reference Katz, Treu and Ayettey50,Reference Lautenschlager and Smith53)} , two in Australia^{(Reference Morgan, Warren and Lubans33,Reference Block, Gibbs and Staiger37)} , one in the UK^{(Reference Revill41)} and one in Spain^{(Reference Perez-Rodrigo and Aranceta47)}.

Of the included studies, ten targeted children aged 7–10 years^{(Reference Miller, Franzen-Castle and Aguirre31,Reference Scherr, Linnell and Dharmar35,Reference Block, Gibbs and Staiger37,Reference Gold, Larson and Tucker38,Reference Revill41–Reference Barnick44,Reference Perez-Rodrigo and Aranceta47,Reference Cunningham-Sabo and Lohse49)} , four studies targeted children aged 11–15 years^{(Reference Gavaravarapu, Saha and Vemula32,Reference Morgan, Warren and Lubans33,Reference McAleese and Rankin40,Reference Hawthorne, Moreland and Griffin48)} and five studies targeted children aged 8–15 years^{(Reference Wolfe and Dollahite34,Reference Beckman and Smith36,Reference Townsend, Johns and Shilts39,Reference Thonney and Bisogni45,46)} .

Seven out of nineteen studies (36·84 %) included parents in the interventions^{(Reference Miller, Franzen-Castle and Aguirre31,Reference Scherr, Linnell and Dharmar35,Reference Block, Gibbs and Staiger37,Reference Gold, Larson and Tucker38,Reference Katz, Katz and Treu42,46,Reference Perez-Rodrigo and Aranceta47)} .

Quality assessment of included studies

The results mainly came from uncontrolled studies and were often based on non-validated outcome measures with no proper adjustment for confounders, which led to the weak global rating for ten studies based on the EPHPP assessment tool⁽²⁹⁾. The quality of nine studies was rated as moderate, and none of the studies were judged as strong.

The data collection method was rated weak for most studies (n 11), largely because there was no information on the measurement instrument’s validity and reliability.

Blinding of students and education providers was generally not possible in the studies. Task outcomes were directly assessed and not likely to be influenced by lack of blinding. Therefore, we assessed blindness as moderate in most studies. The quality assessment of included studies is summarised in Fig. 2 (online Supplementary Table S2).

Strategies and components used in the interventions

Five basic types of strategies were used in interventions aimed at improving FNLIT, including gardening^{(Reference McAleese and Rankin40)}, recipes skill building/cooking^{(Reference Miller, Franzen-Castle and Aguirre31,Reference Revill41,Reference Perez-Rodrigo and Aranceta47)} , food label reading^{(Reference Gavaravarapu, Saha and Vemula32,Reference Katz, Katz and Treu42,Reference Treu, Doughty and Reynolds43,Reference Hawthorne, Moreland and Griffin48,Reference Katz, Treu and Ayettey50)} , food tasting^{(Reference Gold, Larson and Tucker38)} and multi-component interventions^{(Reference Morgan, Warren and Lubans33–Reference Block, Gibbs and Staiger37,Reference Townsend, Johns and Shilts39,Reference Barnick44–46,Reference Cunningham-Sabo and Lohse49,Reference Lautenschlager and Smith53)} . Multi-component interventions included a combination of strategies from gardening/harvesting to food preparation/cooking, recipe skill-building, supporting cultural practices and ethnic foods, food tasting, and food labelling interventions (Table 2).

The variety of skill-building activities introduced by studies is as follows:

Recipe skill building

Two interventions offered recipe skill-building to children and preadolescents (aged 9–15 years)^{(Reference Revill41,Reference Thonney and Bisogni45)} . These included interventions that allowed a child to develop competency in recipe reading. Recipes were purposefully written for children with limited food skills and resources and reflected proper considerations, such as low cost, basic ingredients, basic/simple kitchen equipment, standardised format, numbered preparation steps, core recipes with variations, repetition, and progression of skills, exposure to a variety of foods, dietary guidelines principles, and involved local foods. Workstations were provided for an individual child or a team of two persons with the opportunity to skill-build and gain the confidence to perform the task independently. Working in a small group provided opportunities for peer-to–peer and supportive adult interactions. Additionally, by providing a ‘core’ recipe with simple ingredients, the choice was a practice of the learning experience, allowing youths to make food by their selected ingredients. For example, the ‘create a-flavor’ allowed changes in ‘Apple Cinnamon Toast’ by variations in the type of fruit, bread and seasonings^{(Reference Thonney and Bisogni45)}. In addition, opportunities for conversations about food choices, such as the advantages of whole-grain choices, were provisioned. Overall, these programmes were well received by students.

Food label literacy

Food label literacy interventions were usually part of multi-component school module(s) to promote the skills of use and understanding food label information, as well as informed food choices presented entirely in a one-off session^{(Reference Wolfe and Dollahite34,Reference Katz, Treu and Ayettey50)} or as part of a healthy eating intervention^{(Reference Treu, Doughty and Reynolds43)}. Food label literacy interventions focused on enabling students to (1) explain topics such as nutrients, balanced diets, harmful effects of high fat, sugar and salt foods, and why making informed food choices can benefit their health; (2) recognise deception on packages of food products; (3) identify mandatory information on the labels, aspects they considered while buying packaged foods, defining DV and calculating DV% with differing serving sizes, and key points to make healthy food choices; (4) demonstrate the location of the nutrition facts panel, the ingredient list on food packages, nutrient content declaration (energy, fat, sugar and salt), manufacture, expiration, and best-before dates, and quality symbols; (5) determine foods’ healthfulness according to their labels, nutrition facts panels and the ingredient list on their packages; and (6) grocery store tours^{(Reference Gavaravarapu, Saha and Vemula32,Reference Wolfe and Dollahite34,Reference Treu, Doughty and Reynolds43,Reference Hawthorne, Moreland and Griffin48,Reference Katz, Treu and Ayettey50)} .

Food preparation/cooking classes/clubs

Cooking classrooms and cooking clubs, either embedded into the school curriculum^{(Reference Miller, Franzen-Castle and Aguirre31,Reference Scherr, Linnell and Dharmar35,Reference Block, Gibbs and Staiger37,Reference Gold, Larson and Tucker38,Reference Barnick44,Reference Thonney and Bisogni45,Reference Cunningham-Sabo and Lohse49,Reference Lautenschlager and Smith53)} or delivered in the form of an after-school food club^{(Reference Wolfe and Dollahite34,Reference Townsend, Johns and Shilts39,Reference Revill41,Reference Perez-Rodrigo and Aranceta47)} , are another approach used to promote food skills (Table 3). Food clubs were held over several weeks (e.g. 20 weeks). The length of cooking classes or courses varied from a few hours per week to multiple days of training. Classroom-based activities focused on science, health, nutrition, literature and field trips to grocery stores, restaurants, nature centres and cultural events. Children were involved in an in-depth demonstration focused on specific food/ethnic foods^{(Reference Lautenschlager and Smith53)} or skills, such as preparing delicious foods, identifying food safety and self-efficacy. Students became more confident and independent by learning the importance of healthy nutrition and hands-on skills in a kitchen setting. In the interactive cooking classes, students cook along with a chef and their peers in real time. Designed to look and feel like they were cooking in their own home, each student had his/her own cooking station, complete with sinks, aprons and cookware sets. Interventions offered hands-on skills, along with food-knowledge building^{(Reference Miller, Franzen-Castle and Aguirre31,Reference Block, Gibbs and Staiger37,Reference Townsend, Johns and Shilts39,Reference Revill41,Reference Barnick44,Reference Perez-Rodrigo and Aranceta47,Reference Lautenschlager and Smith53)} . One intervention used cooking demonstrations using the ‘Cooking Up Healthy Choices’ curriculum. Cooking Up Healthy Choices was a series of five cooking demonstration sessions that allowed students to get familiar with a variety of vegetables, observe cooking methods, understand related nutrition concepts and experience the preparation of recipes using all five senses^{(Reference Scherr, Linnell and Dharmar35)}.

Food tasting

Students participated in communal food activities that impacted food knowledge and fostered positive food nature^{(Reference Morgan, Warren and Lubans33,Reference Wolfe and Dollahite34,Reference Gold, Larson and Tucker38,Reference Barnick44)} . Students brought new food to the class and talked with each other about how they tasted. They were encouraged to notice and enjoy the sensory characteristics of food and eagerly shared their pleasure with their peers. In the ‘Cooking With Kids (CWK)’ intervention, students were exposed to tasting lessons^{(Reference Cunningham-Sabo and Lohse49)}. Through these sessions, students would learn to try new food as one of the components of functional skills of FNLIT (Table 3).

Gardening/harvesting

Seven studies specifically focused on gardening/harvesting interventions^{(Reference Morgan, Warren and Lubans33,Reference Scherr, Linnell and Dharmar35,Reference Block, Gibbs and Staiger37,Reference McAleese and Rankin40,Reference Barnick44,46,Reference Lautenschlager and Smith53)} . These programmes were carried out as gardening lessons in the classroom curriculum. Children assigned to gardening groups received weekly lessons focused on garden activities and the food system. They were engaged in either doing crafts (e.g. photography) or gardening in the afternoons. Volunteer adolescents sold their planted products in the farmers’ market during the weekends. They also went to the ‘Farm Camp’ and learned how a small-scale, organic, cooperative farm operates.^{(Reference Lautenschlager and Smith53)}.

Supporting cultural practices and ethnic foods

Some programmes consisted of strategies to increase children’s willingness and cognition towards ethnic and indigenous foods. Students were introduced to an ethnic meal prepared by young cooks in this programme and tried ethnic and unfamiliar foods^{(46,Reference Lautenschlager and Smith53)} . Understanding diverse ethnic and cultural practices related to meal preparation and consumption is one layer of FL^{(Reference Vidgen and Gallegos2)}.

Implementation methods of the interventions

The educational/training sessions were presented mainly by lectures, pictorial booklets, and posters, accompanied by power points, videos, and short animation films to engage, motivate and inform the students. Also, some group activities were performed, for example, assigning teams of students to search through a grocery bag containing food products, such as cereals, crackers, or snack bars, and decide which products are healthful ‘clued-in’ and which are less healthy ‘clue-less’^{(Reference Katz, Treu and Ayettey50)}. Other teachings and learning activities included take-home challenges and parents’ newsletter, role-playing, playing together, grocery store tours, hands-on activities, doing crafts (photography) and animation film for entertainment education.

Delivery formats of interventions in the fifteen of the nineteen studies (78·94 %) were by teachers^{(Reference Miller, Franzen-Castle and Aguirre31–Reference Morgan, Warren and Lubans33,Reference Scherr, Linnell and Dharmar35,Reference Block, Gibbs and Staiger37–Reference Barnick44,46,Reference Perez-Rodrigo and Aranceta47,Reference Cunningham-Sabo and Lohse49)} . Investigators supplemented information only when it was necessary. Some other interventions (n 4) were facilitated by community health educators^{(Reference Wolfe and Dollahite34)}, registered dieticians^{(Reference Hawthorne, Moreland and Griffin48)}, as well as community members involved in the programme^{(Reference Thonney and Bisogni45)}.

Interventions in the promotion of Food and Nutrition Literacy dimensions

Effectiveness of interventions

Because of the low quality of the studies, we can draw no firm conclusions regarding the effective components of food/nutrition literacy interventions. However, the following common factors were noted within the interventions successful in more than one dimension of FNLIT, especially interactive and critical aspects, which were identified as promising. Four out of nineteen studies included the following factors (21·05 %):

• the interventions which tailored their activities and presented information to the needs and interests of students^{(Reference Block, Gibbs and Staiger37)};

• the interventions that were incorporated into the existing curriculum and facilitated by teachers^{(Reference Miller, Franzen-Castle and Aguirre31,Reference Block, Gibbs and Staiger37,46)} ;

• interventions mainly used promising strategies/methods, including pleasurable hands-on food education, school gardening programmes, kitchen classrooms, family cooking, eating and playing together, and supporting cultural practices and ethnic foods^{(Reference Miller, Franzen-Castle and Aguirre31,Reference Block, Gibbs and Staiger37,46,Reference Hawthorne, Moreland and Griffin48)} that led to improvements in functional, partly interactive and critical skills.