(1) Eligibility criteria

The emphasis of this review is largely placed on the intervention aspect of food literacy. Therefore, maximum data related to intervention will be extracted. In doing so, those studied will only be included which are (1) peer-reviewed, (2) studied on humans, (3) studied in high-income countries, (4) described a food literacy intervention implemented on adults aged 18 years or above, (6) published from 2001 to 2022 (7) published or translated in English.

(2) Search strategy

To visualize the search plan, four main theoretical constructs related to the research question were identified first (presented in Fig. 1).

Fig. 1

Visual presentation of search adaptation

As recommended in all JBI types of reviews, a three-step search strategy was developed by all three authors, along with the consultation of an academic librarian.

Step 1

On July 5, 2023, an initial search was carried out on MEDLINE (via EBSCOhost), Cochrane Library, and OSFHOME (Open Science Framework) databases using the keywords “food literacy,” “intervention,” “adults,” “healthy eating,” and “socioeconomic position. No systematic review was found to be registered on any platform that sufficiently addressed the research question of this study. However, the initial search also revealed significant limitations in the search strategy.

Many articles in the database used these search terms in a different context than intended. For example, the phrase “healthy eating” has multiple meanings and has been used in various contexts. In addition, no empirical study has explored the connection between food literacy and socioeconomic position till now. As a result, relying solely on the above-mentioned keywords either failed to yield related materials that did not explicitly use the search term or returned irrelevant materials.

Step 2

As such, the search strategy was adjusted to include only the keywords “food literacy” “intervention” and “adults”, along with the index terms used to describe these three constructs as identified in the titles/abstracts of articles from the initial search. The extraction of socioeconomic factors and any indication of healthy eating behavior (dietary behavior) will be carried out manually. A trial of search, using the preferred keywords and the index terms used in each database is shown in Appendix 1.

The second search using the modified search strategy will be run in three electronic databases (MEDLINE (via EBSCOhost), Scopus, and CINHAL) by AM in November 2023.

Step 3

To ensure the completeness of the search process, both forward and backward citation searches will be performed(QUT Library Guide, 2023).

After each search, all identified citations will be gathered and uploaded to the referencing software EndNote 20. Duplicates will then be removed before exporting the citations to Covidence, a screening and data extraction tool will be used in systematic reviews (https://www.covidence.org/).

(3) Study selectionIntegrating different types of studies

The review will encompass diverse types of studies to gain a better understanding of multifaceted phenomena. These will include quantitative studies, which measure the effects of food literacy interventions, qualitative studies that focus on the experiences of those who attended any food literacy program, and mixed methods studies that combine both quantitative and qualitative approaches.

While reviewing quantitative studies, trials of food literacy programs/interventions that aimed to promote food literacy behavior and healthy eating will be looked at. The analysis for the trials will involve a thorough examination of the pre- and post-data on the outcomes that were reported, along with that of a comparable control or comparison group.

The aim of reviewing qualitative studies is to explore adults' perspectives and experiences of attending food literacy programs. The focus is on identifying what attendees have reported experiencing as a result of participating in these programs. Initial reviews of the available literature indicate that attendees of such programs have reported positive changes in their food habits, including eating more fruits and vegetables, gaining confidence in cooking, using healthier ingredients, adopting appropriate cooking methods, substituting ingredients with less expensive options, making informed decisions when selecting food items, and stretching their meal budgets [13, 76].

Process of selection

All authors (AM/HV/DG) will conduct a pilot test by screening the title and abstract of 10% of the articles randomly selected from the pool of the saved articles in Covidence against the inclusion criteria. Once a consensus is reached, the first author (AM) will screen the title and abstract of the remaining articles. The exact process will be followed for assessing articles in full text. After reaching an agreement, the first author will retrieve the full text of the primarily selected citations and assess them in detail against all the inclusion criteria, including language, participants, geography, and intervention. Another reviewer (HV) will repeat this process independently. Any disputes will be resolved by consensus or with the involvement of a third reviewer. Finally, the scoping review will include all the publications that meet the eligibility criteria.

(4) Evaluation of food literacy intervention

Although previous systematic reviews of food literacy interventions expose the inadequacy of the evaluation method, it is still crucial to assess the effectiveness of the interventions through post-program follow-up evaluations [9]. It is also important to select an appropriate evaluation design that corresponds to the level of development when assessing a program. This review will follow three main types of evaluation methods: process, outcome, and impact evaluation [32, 55].

The method of process evaluation is used to determine if program activities have been executed according to plan and if they have resulted in specific outputs [32]. The relationship between impact and outcome can be explained as follows: outcome is the goal of any project (intervention), while impact is the objective. To clarify, the outcome is characterized by the desired changes in targeted health behavior that are sustained over a long period of time. Impact evaluation provides information about the observed changes or “impacts” produced by the intervention [32]. For instance, when implementing a food literacy program for adults, the objective is to improve their food literacy behavior, resulting in a sustained improvement in their dietary behavior, which is the ultimate goal or outcome [5, 6]. As such, for this scoping review, the impact is the modification in food literacy behavior, and the outcome is the modification in eating behavior. How the data for process, impact, and outcome evaluation will be extracted is described in the coming paragraphs.

Furthermore, the assessment method is consistent with Vidgen’s ([72], p. 75) “second model of food literacy”, which is illustrated in Fig. 2 below. This model not only illustrates the connection between food literacy and nutrition but also provides guidance for process, outcome, and impact evaluations in an ideal food literacy intervention. The insights gained from these evaluations can be applied to improve the development and execution of future interventions [72], p. 81). Hence, this model will guide the evaluation process.

Fig. 2

Second model of food literacy, depicting the relation between food literacy and nutrition. Note: Adapted from Food Literacy: What Is It And Does It Influence What We Eat? by [72], p. 75

Below is an outline of how three types of evaluation will be implemented when reviewing various interventions.

Process/implementation evaluation

The 11 components of food literacy may serve as a framework for the process evaluation of a food literacy intervention, as suggested by [72], p. 81. Some of the constructs that will be investigated under the “process evaluation category” are which components of food literacy were addressed in the program, how the programs were designed, the percentage of adult participants, the records of their socioeconomic and sociodemographic characteristics, if the program’s effectiveness was measured according to attendee feedback, and what were the barriers/facilitators to implementation of program activities (Table 2).

Table 2 A list of framed questions will help in the process of evaluationImpact evaluation

To ensure food literacy is held accountable for driving healthy eating practices, it is essential to measure the impact of food literacy intervention on health outcomes [27]. In regard to that, Vidgen [72] proposed analyzing the constructs of “certainty”, “choice”, and “pleasure” (Fig. 2) is crucial in determining the impact of a food literacy intervention. For instance, a food literacy program can have a positive “impact” by increasing “pleasure” in cooking or by providing more “choices” in selecting healthy and affordable food from the local food environment. Therefore, to evaluate the impact of food literacy intervention, this paper will gather data on changes reported in various components of food literacy. These components include planning food intake (under the domain of planning, in component 1.1), reducing consumption of fast food and sugary drinks (under the domain of eating, in component 4.2), and increasing self-reported cooking skills (under the domain of preparation, in component 3.1).

These outcomes are indicative of the successful implementation of food literacy intervention and can guide future development [6] (Table 3).

Table 3 A set of questionnaires, which will be used to extract data for impact evaluationOutcome evaluation

This review aims to determine any enduring effects on eating/dietary behavior after the delivery of a food literacy program as “outcome evaluation”. In a prior study [5], a comparable methodology was employed to assess the impact of food literacy by tracking modification in dietary behavior, which was deemed a critical metric for measuring outcomes of food literacy interventions. Next, there is a discussion of what is meant by “eating behavior” and the method that will be utilized to track any changes in such behavior.

The concepts of “eating behavior” and its related terms, including “dietary behavior”, “dietary intake”, “eating habits”, “diet”, and “food choice”, are broad and ambiguous ideas, and these terms are used interchangeably in various academic fields [42]. In general, the term ‘eating behavior” or “dietary behavior” is a conclusive idea that encompasses all the factors related to food consumption, including diet quality, food preferences and motives, eating patterns, and diet-related chronic diseases [37, 60]. For this paper, the term “eating behavior” will be used consistently to refer to all the above-mentioned concepts.

There are different ways to measure different aspects of healthy eating behavior. In the field of dietary behavior research, self-reported measures, such as 24-h dietary recalls, food records/diaries, and food frequency questionnaires (FFQ), are commonly employed to collect data [53]. This is because it is generally not possible to objectively assess the usual dietary intake in community-dwelling individuals [35]. Hence, as measures for dietary behavior, this review will include previous studies that have reported dietary outcomes through self-reported measures along with other measures.

This review will rely on a range of measures, including the following:

Measures of diet quality: Diet Quality Indices (DQIs) serves as tools for assessing an individual’s overall diet quality. These scores food and/or nutrition intakes and sometimes lifestyle factors based on how closely they align with dietary guidelines [77]. Examples of DQIs are the Healthy Eating Index (HEI), the Diet Quality Index (DQI), the Healthy Diet Indicator (HDI), the Mediterranean Diet Score (MDS [25, 75], and Single-item self-rated diet measure (SRD) [23].

Dietary intake indicator: e.g., the Household Dietary Diversity Score (HDDS; measures food accessibility and socioeconomic status based on types and quantity of food consumed in 24 h [34] (Table 4).

Table 4 Some structured questions, that will help to extract data for outcome evaluation (5) Data extraction

The lead researcher will extract the content of each study independently. The extracted findings then be shared with the supervisory team for approval. In the event of any conflicts, they will be resolved through discussion. The data extraction matrix will be revised and may be modified if required during the process.

Following the protocol, the data extraction matrix (an Excel sheet) will summarise the data under four main headings: (1) description of studies, (2) process evaluation, (3) impact evaluation, and (4) outcome evaluation. Under these four headings, all the single constructs will be assessed. Those constructs are listed below. An example of a data extraction matrix is attached in Appendix 3.

Study details: (1) author, (2) study location, (3) sample size, (4) study design, (5) theoretical framework applied, (6) year of publication of the study results, (7) published in a journal

Population details : (8) socioeconomic characteristics of the target group (high and low socioeconomic group, description of socioeconomic factors, such as income, education, occupation), (9) sociodemographic characteristics of the target group

Intervention details (10) Name of the intervention (food literacy program), (11) components of the intervention (e.g., the components of food literacy addressed), (11) duration of the interventions, (12) measurement tool (e,g., food literacy scale, food literacy questionnaire, & FFQ),

Impact details: (13) report of changes in food literacy behaviour, (14) measurement tool, (15) findings

Outcome details: (16) report of changes in dietary behaviour, (17) measurement tool, (18) findings

(6)Assessment of risk bias: study appraisal

To evaluate the quantitative aspects of the articles, the JBI critical appraisal checklist for analytical cross-sectional studies, which is an eight-item questionnaire, will be employed (The Joanna Briggs [63]. Meanwhile, for the qualitative studies, the ‘Consolidated Criteria for Reporting Qualitative Studies (COREQ), which consists of 23 questions, will be used [67].

(7) Analyse and synthesize the evidence Synthesis of qualitative papers

After extracting the data, the information extracted from each paper, including study details, population details, intervention, impact, and outcome details, will be utilized to create evidence tables providing an overall description of the included studies.

Subsequently, two team members (AM and HV) will independently analyze the extracted data based on those predetermined categories.

Qualitative papers will be subjected to thematic analysis, as described by Braun and Clarke [11]. The thematic analysis aims to identify significant data patterns (“themes”) and establish a visual network and conceptual connections among these themes to address the primary and secondary research questions specific to this systematic review. During this process, both reviewers (AM and HV) will independently conduct line-by-line coding from the findings of the selected studies to identify recurring, unique, and contradictory content. These codes will then be utilized to form themes and a series of sub-themes [65]. The reviewers will utilize computer-assisted qualitative data analysis software (CAQDAS) such as NVivo to assist in this step. While the researcher creates the codes, NVivo can help with sorting, labeling, and organizing the codes (referred to as “nodes” in NVivo) and the data (Dhakal, 2022,NVivo, 2023). As thematic analysis is a comprehensive process, the reviewing team will convene for multiple meetings to arrive at consensus decisions. Investigator triangulation will be employed during this process, with two or more researchers involved, to mitigate personal bias and ensure the inclusion of diverse perspectives [15].

Synthesis of quantitative papers

Due to the inherent nature of systematic reviews, it is anticipated that this systematic review will encompass a wide range of quantitative studies characterized by diversity in the intervention (including duration and delivery model), study design (e.g., cross-sectional and longitudinal cohort), study participants (e.g., physical condition, age, gender, and location), and the outcomes/effects (varied measurement methods and durations). This variability is commonly referred to as "heterogeneity" in research [17, 31]. As heterogeneity is expected, this review will use a meta-analytical method to combine study estimates and obtain a summary estimate(e.g., mean differences, standardized mean difference, and its 95% confidence interval) [54]. The most appropriate approach for the meta-analysis in this case is a random-effects meta-analysis, which will effectively assess the variations in the effects of different interventions [54]. In addition, Forest plots will be used for visual examination of heterogeneity [31]. To assess the degree of heterogeneity statistically, three measures will be employed: (1) Cochran’s Q to evaluate whether the proportion of successes is consistent across groups, (2) Higgin’s and Thompson’s I2 to assess the percentage of variability in effect sizes not caused by sampling error [30], and (3) Tau-squared to estimate the variance of the underlying distribution of true effect sizes [29].

The results from both the quantitative and qualitative synthesis will be integrated to produce the final synthesis that will help gain a comprehensive understanding of how different aspects of the research relate to one another. The qualitative papers will be analyzed to develop a set of recommendations for interventions that are in line with the perspectives of adult attendees. These recommendations will then be utilized to evaluate the interventions analyzed through quantitative synthesis to determine the level of alignment between the interventions and our recommendations [28, 52].

(8) Report the findings

The findings of this review are intended for publication in a scholarly journal that focuses on public health or nutrition science. Additionally, the result may be shared through other networks, such as conferences. As a part of an effort to ensure data transparency and accessibility, all data resulting from this review will be uploaded to the Queensland University of Technology’s repository. The reviewers wish for the significant findings to be widely and readily available to those who can benefit from this research.

(9) Strengths and limitations

This will be the first review to synthesize evidence on the link between food literacy and socioeconomic position and healthy eating.

The results will aid in comprehending whether previous food literacy interventions have effectively assisted individuals belonging to low-socioeconomic groups in adopting healthy eating habits.

There is a lack of studies that have evaluated post-program analysis of food literacy intervention, specifically in relation to the food literacy domains or the three levels of food literacy outcome.

It is possible that some interventions aimed at improving food literacy behavior may be missed due to the fact that not all studies use the term “food literacy” directly and instead focus on enhancing specific components related to it.

It is important to note that the review will have some limitations regarding bias. Specifically, certain countries, papers written in languages other than English, and specific population groups were intentionally excluded. As a result, the selection process was significantly biased. The decisions have been taken to make sure that the review's scope is narrowed down and relevant information is gathered. In the upcoming reviews, it would be advantageous to examine literature from low- to middle-income nations and also to involve children and elderly individuals who have firsthand experience with attending a food literacy program.