From Policy to School Plate: Assessing the Translation of Aquatic Food Recommendations into School Meal Systems in Low- and Middle-Income Countries

1. Introduction

School meal programs reach over 466 million children annually and support nutrition, education and local food systems (GCNF 2024b; World Food Programme (WFP) 2024). Governments increasingly seek alignment with dietary guidance, sustainability, climate resilience and local procurement (FAO 2019b; WHO 2026), but implementation depends on governance, budgets, procurement, supply chains, infrastructure, food safety and local availability and preferences (Micha et al., 2018; Cohen et al., 2021; WHO 2026).

Poor diet quality remains a major global health and nutrition challenge, characterized in many populations by inadequate intake of both high-quality protein and essential micronutrients. More than half of the world's population consumes inadequate amounts of at least one of several key micronutrients, including iron, zinc, calcium, iodine, folate, and vitamins A, B12 and C (Passarelli et al., 2024). Undernutrition also remains widespread: in 2024, an estimated 23.2% of children under five years of age—150.2 million children—were affected by stunting (UNICEF, WHO and World Bank, 2025). Inadequate dietary intake of protein and essential micronutrients, as well as long-chain omega-3 fatty acids, can constrain physical growth, development and cognitive function, particularly during periods of rapid growth, which compromises growth, learning and long-term health (Nyaradi et al., 2013). Maintaining nutritional gains from the first 1,000 days requires attention throughout childhood and adolescence—the subsequent “7,000 days” (Yakubu et al., 2026; WFP, 2024), which encompass growth, puberty, brain development and lasting dietary preferences (Hanley-Cook et al., 2024; WFP 2024). Calcium requirements increase during adolescence, while girls have high iron requirements associated with menstrual losses (Matkovic 1992; Low et al., 2016).

Aquatic animal-source foods (ASFs), including fish, shellfish, molluscs and other aquatic animals, are rich in bioavailable micronutrients (e.g., iron, zinc, vitamin B12 and iodine), long-chain omega-3 fatty acids, including docosahexaenoic acid (DHA) and eicosapentaenoic acid (EPA), and high-quality protein that support child and adolescent growth and cognitive development (UN Nutrition 2021; FAO/WHO 2024a; O’Connor et al., 2025; Consalez et al., 2022). In many low- and middle-income countries (LMICs), small fish and other locally available aquatic ASFs can be relatively affordable and nutrient-rich, particularly when they are consumed whole, including the bones and organs (Johnsson et al., 2025; Robinson et al., 2022, 2025). Seaweed and other algae provide iodine, fibre and bioactive compounds; some also contain vitamin B12, with variable bioavailability (Koseki et al., 2023). Nutrient composition varies by species and growing conditions (Park et al., 2023), and a single portion of some products may exceed the tolerable upper iodine intake level (Aakre et al., 2021), with potential adverse health effects (Blikra et al., 2024).

Fish and other aquatic food consumption among many children and adolescents remains below recommendations. In high-income countries, barriers include household dietary practices, cost, sensory preferences, perceived quality, contaminant and sustainability concerns, and limited preparation confidence (Ansai et al., 2025; Derbyshire 2019; Utri-Khodadady et al., 2024). Evidence from LMICs, largely focused on infants and young children, additionally identifies limited access, seasonal food insecurity, delayed introduction during complementary feeding, concerns about bones, caregiver beliefs, cultural norms and intra-household food allocation. These barriers can persist even in fish-producing communities (Bogard et al., 2017; Byrd et al., 2021; Thorne-Lyman et al., 2017). School meals may improve equitable access and have potential to support diet quality (Kante et al., 2026; Lehner et al., 2020).

Food-based dietary guidelines (FBDGs) provide population-level recommendations for healthy, culturally appropriate diets and can inform institutional guidance (FAO 2019b; Herforth et al., 2019). School food policies shape foods provided, sold or marketed in and around schools (FAO 2019a; WHO 2026). Nutrition guidelines and standards for school meals translate dietary objectives into menu-planning, procurement and provision guidance (FAO 2019b; WHO 2021; WHO 2026), while menus and recipes specify planned provision. Here, ‘school food’ refers to the broader policy and food environment, and ‘school meals’ to organized meal provision. Countries may have school food policies without national school meal programs, or school meal programs without comprehensive school food policies.

This study examined aquatic foods across the school food policy landscape in LMICs. We aimed to: (1) identify aquatic food references in FBDGs, school food guidance, menus and recipes; (2) map document availability and aquatic food inclusion across the policy-to-school-plate pathway; (3) assess stakeholder-reported planning, provision and implementation barriers; and (4) identify opportunities for feasible, context-appropriate inclusion of aquatic foods in school meals.

Our conceptual framework (Figure 1) traces dietary recommendations through school food guidance, menus and recipes, foods served, and children's dietary exposure, health and education outcomes. School meals can also support livelihoods and wider food system outcomes. We use aquatic foods to identify discontinuities across these stages. To our knowledge, no previous multi-country study has systematically mapped aquatic foods across these layers.

2. Methods

We included 132 LMICs based on the World Bank 2026–2027 classification, retaining Costa Rica despite its recent transition to high-income status (World Bank 2026). Document availability was assessed independently of reported school meal program status: policies may precede national programs or support subnational, community-led or future provision. Restriction to reported national programs would rely on incomplete, time-varying program-status data and exclude countries with undocumented implementation.

2.1. Document Search and Review

Inclusion criteria, sources, search period, search terms, numbers of records screened and excluded, and data extracted from each document are summarized in Table 1. Searches were conducted from 3 November 2025 to 15 March 2026, and eligible documents were included regardless of publication date. When multiple documents of the same type were identified for a country, the most recent was included.

‘School food guidance documents’ includes policies, standards and practical tools governing the nutritional quality and safety of foods provided or sold in and around schools: school meal nutrition guidelines and standards, menu-planning guides, vendor regulations, and recommended, restricted or prohibited food lists.

All available daily, weekly or monthly menu cycles were reviewed. For recipe books containing more than 30 recipes, ChatGPT 5.6 Sol Light (OpenAI; last accessed 17 September 2026) screened each recipe against predefined aquatic food inclusion and type criteria. The model was supplied with a standardized prompt and output template to record country, total recipes, recipes containing aquatic foods and aquatic food types. A researcher independently reviewed a random sample of 3 of 26 recipe documents (12%) using the same criteria; discrepancies were corrected before finalizing the dataset. Table 1 summarizes records and countries reviewed; individual recipe counts are reported in the Results.

2.2. Stakeholder Survey

The online survey was administered in English, French and Spanish using non-probability convenience and snowball sampling. Professional contacts received the link through LinkedIn and targeted distribution lists and were invited to circulate it further.

Target respondents included government officials, non-governmental and international organizations, implementers and suppliers involved in school meals. Responses could concern national, subnational, community or informal initiatives. The survey collected quantitative and qualitative information on aquatic foods in menus, recipes and meals served; barriers and enablers rated on a 4-point Likert scale from not important (1) to very important (4); and supporting documents for review (Section 2.1). Most survey questions were presented as multiple-choice items, allowing respondents to select one or more response options. Conditional branching distinguished suppliers or producers from other stakeholders. The survey was open from 3 April to 8 June 2026.

Clearance for the treatment of personal data was obtained from SIKT (Norwegian Agency for Shared Services in Education and Research) on 4 March 2026 (Notification Form 999356), and survey respondents were given comprehensive information on the use of their data and then asked to actively consent to participate in the online survey.

2.3. Analysis

Documents and survey responses were summarized descriptively. Country-level mapping and respondent-level analyses were treated separately because the convenience sample was not representative. Document discontinuity refers to reduced specificity, coverage or implementation between successive pathway stages, including absent aquatic foods or insufficient frequency or quantity relative to dietary objectives. We compared document availability and aquatic food content across stages and assessed planned versus reportedly served foods within paired survey responses. These comparisons did not establish causal mechanisms or verify national implementation. Comparable portions reported in grams were assessed against the Global Diet Quality Score (GDQS) Meal and Menu Metrics fish and seafood thresholds for children aged 5–9 and 10–14 years: >40 g indicates a high contribution to meal quality, 8–40 g medium and <8 g low (Intake 2025).

Survey questions covering the same food category at different stages of the pathway were paired a priori to compare planned and reported provision. Respondent-level agreement was assessed using Cohen's kappa: 0.01–0.20 slight, 0.21–0.40 fair, 0.4–0.60 moderate, 0.61–0.80 substantial and >0.80 almost perfect. Multiple respondents could represent one country; estimates were not nationally representative. Analyses used R 4.5.1 and RStudio 2025.09.1+401. Responses to one open-ended survey question on recommendations for improving the inclusion of aquatic foods in school meals were analyzed thematically by two reviewers. The reviewers familiarized themselves with the data and iteratively generated inductive codes, which were grouped into broader themes following Braun and Clarke’s six-phase framework (Ahmed et al., 2025; Campbell et al., 2021). Themes included policy development, procurement reform, improved infrastructure, local sourcing, food safety systems, product innovation, and capacity strengthening.

Data from the Global Survey of School Meal Programs, conducted by the Global Child Nutrition Foundation (GCNF) in 2019, 2021, and 2024, were used to contextualize findings from the document review and stakeholder survey in relation to countries reporting large-scale school meal programs (GCNF, 2024a).

3. Results

3.1. Availability of Policy and Implementation Documents

At least one relevant document type was identified for 101 of 132 countries (76%), but documentation availability declined downstream (Table 2; Supplementary Table S1). Among the 132 countries, 94 reported school meals in at least one GCNF survey (2019, 2021 or 2024), 22 never responded, and 16 reported no large-scale program at the time. Reported programs included national, sub-national, community/parent-led provision in Uganda and Tanzania and emergency provision in Yemen. Among the 31 countries without identified documents, 16 reported school meal programs, nine never submitted GCNF survey responses, and six reported no program (GCNF 2024a).

FBDGs were identified for 63 countries (48%). Of the 63 countries with FBDGs, 8% were low-income, 35% lower-middle-income and 56% upper-middle-income countries. They were most often endorsed by ministries of health (n=49), agriculture (n=5), or both (n=2). Seven national FBDGs included recommendations for school-aged children, generally within a broader category spanning children and adolescents; Bolivia and Bhutan had separate guidance documents for school-aged children or adolescents. School food guidance was identified for 71 countries (54%): 10% low-income, 39% lower-middle-income, 49% upper-middle-income, plus Costa Rica. Of 28 school nutrition guidelines and standards (NGS), issuers most often included ministries of education (n=9), health (n=6), or both ministries (n=4). Mongolia had both NGS and a practical guide. India had ministry of health-led nutrient standards and education-led food standards and food education activities.

Eighty-one menu or recipe documents represented 57 countries (43% of the 132 in our search): menus alone were identified for 31 countries, recipes alone for ten, and both document types for 16. Menus were therefore available for 47 countries and recipe documents for 26. Only 11% of countries (n=14) had accessible documentation across all stages of the pathway, including FBDGs, school food guidance, menus, and recipes. Conversely, no accessible documentation was identified for 23% of countries (n=31).

3.2. Inclusion of Aquatic Foods Across the Policy-to-School-Plate Pathway

Aquatic foods were explicitly included in 62 FBDGs (98% of identified FBDGs; 47% of all 132 countries). One FBDG was not fully accessible and could not be assessed for aquatic food inclusion. Aquatic foods were usually represented as finfish within broad animal-source or protein-rich food groups (Figure 2). Seaweed and other algae were included in only five FBDGs: the Dominican Republic, Ecuador, India, Indonesia and the Philippines. Where included, they were grouped with fish or mentioned in relation to specific nutrients, such as iodine or omega-3 fatty acids.

Among the 62 accessible FBDGs that included aquatic foods, 29 provided quantitative fish portion sizes and 11 provided semi-quantitative descriptions, such as a ‘palm-sized’ portion (Figure 3). Twelve specified a fish-specific consumption frequency. Quantitative recommendations for non-fish aquatic foods, such as molluscs or crustaceans, were not identified.

Among 71 countries with school food guidance documents, 75% (n=53), or 40% of the total 132 countries included in our review, explicitly referenced fish or other aquatic foods (only China referenced seaweed). Documents that did not mention fish were typically nutrient-based or referred only to animal-source foods more generally. Two countries (Algeria and Jordan) categorized fish along with meat, which was restricted in school canteens or subject to veterinary certification requirements. Approximately half of the school food guidance documents reviewed included serving size guidance (n=28) or frequency recommendations for aquatic foods (n=28), while only 11 provided both frequency and serving size guidance. The mean frequency recommendation and serving size recommendation range is reported in Figure 3. No quantitative recommendations for non-fish aquatic foods were included in school food guidance documents.

Aquatic foods appeared in menus and/or recipes for 32 countries (24% of 132 countries; or 56% of countries for which we found menus and/or recipes). Aquatic foods were included in both menus and recipes in 12 countries (9% of all 132 countries), in menus only in 14 (11%), and in recipes only in six (5%). Among countries with available documents, aquatic foods appeared in menus in 26 of 47 countries (55%) and in recipes in 18 of 26 countries (69%). Three countries included seaweed in recipes – Mongolia, Timor-Leste and Ukraine. Figure 3 summarizes quantifiable serving sizes from recipes and serving frequencies depicted in menus. Among 1,502 individual recipes reviewed, 7% (n=101) included at least one aquatic food, with fresh or frozen fish dominating 61% of individual recipes that included aquatic foods (n=62). Figure 4 details the inclusion of aquatic foods in individual recipes across countries where recipes were reviewed (n=26); percentages reflect the documents reviewed.

Taken together, aquatic food inclusion was 98% among accessible FBDGs and 75% among school food guidance documents. Within the operational-document stage, aquatic foods appeared in 55% of assessed menus and 69% of assessed recipe documents, and 7% of the 1,502 individual recipes reviewed. Denominators differ by stage and describe accessible evidence. Among the 14 countries where the complete policy-to-school-plate pathway could be assessed, ten countries (Belize, Mexico, Moldova, Mongolia, Paraguay, Saint Lucia, South Africa, Sri Lanka, Timor-Leste, Vietnam) demonstrated inclusion of aquatic foods across FBDGs, school food guidance, menus, and recipes. Figure 5 presents a high-level overview of the inclusion of aquatic foods across the policy-to-school-plate pathway. Supplementary table S1 provides more detail on available documents and inclusion of aquatic foods in each.

3.3. Stakeholder Survey Findings

The stakeholder survey generated 46 complete responses from 23 countries, on average two respondents per country, although representation was unequal., Most responses came from lower-middle-income countries (65%), followed by upper-middle-income countries (20%), then Costa Rica (9%) and Malawi (4%) as the only high- and low-income countries that responded to the survey. Respondents primarily represented international organizations (35%) and government institutions (24%) and reported working at national (80%), provincial or regional (41%), community (35%) or international (30%) levels (Supplementary Table S3). Three-quarters of respondents reported school nutrition guidelines or standards and 71% reported official school menus or recipe books. Out of the 23 countries represented in stakeholder survey responses, all reported school meal programs in past GCNF surveys, except for four: two countries (Iran and Fiji) have never responded to a GCNF survey, whereas the other two countries (Samoa and Solomon Islands) have responded and noted that there is no large-scale school meal program in the country (GCNF 2024a).

Paired responses were available from 30 respondents representing 16 countries. Overall agreement between foods included in official menus or recipes and foods reported as generally served was substantial across all food groups (κ = 0.71). For aquatic animal-source foods (fish and other aquatic animals), 80% reported inclusion in official menus or recipes, compared with 57% reporting that they were generally served; agreement was moderate (κ = 0.49). Seaweed and other algae appeared in menus or recipes for 20% of respondents but were reported as generally served by 7%, with slight agreement (κ = 0.17) (Supplementary Table S4). Because observations were respondent-level and multiple respondents could represent one country, these findings indicate perceived alignment rather than nationally representative implementation gaps.

Aquatic food procurement was predominantly reported from local fish markets or small-scale traders (73%). Fresh aquatic foods were the most frequently reported form in menus or recipes (79%), followed by frozen (42%), dried (38%), canned (33%) and fish powder (17%) (Supplementary Table S5). Freshwater fish, large marine fish and small fish were commonly reported; however, the questionnaire examples did not consistently distinguish small marine from small freshwater species, so these categories should be interpreted cautiously. Small fish were reported in menus or recipes by 58% of respondents and as generally served in school meals by 77%.

Respondents rated food safety and quality and affordability most highly among enabling factors (both 3.82/4). Cost and inadequate funding received the highest barrier score (3.77/4), followed by infrastructure limitations, safety concerns and unstable supply (Supplementary Table S6). Open-ended responses additionally mentioned centralized procurement, limited monitoring, weak policy prioritization, storage constraints and cultural considerations as barriers. These respondent-reported perceptions were not designed to establish regional prevalence or causal importance.

3.4. The Policy-to-Plate Pathway

Table 3 provides selected examples of aquatic food portions and frequencies in FBDGs, school food guidance, menus and recipes with stakeholder-reported provision, including countries with survey evidence of provision but incomplete upstream documentation. Further examples are available in Supplementary Table S2. Quantitative policy alignment was evident in Malaysia and South Africa, and alignment between school food guidance and menus or recipes in Moldova. Timor-Leste, Sri Lanka and Paraguay showed alignment across documented stages, although Paraguay had no survey evidence of provision. Most surveyed countries had broad protein or animal-source food categories in guidance (Bhutan, Cambodia, Costa Rica) or lacked identified upstream guidance (Lao PDR, Marshall Islands).

In several countries, recommended serving sizes in school food guidance documents exceeded the GDQS Meal and Menu Metrics threshold for a high contribution to meal quality (>40 g of fish or shellfish) (Bhutan, Costa Rica, Samoa and Timor-Leste). Timor-Leste and Sri Lanka had reported serving sizes in menus or recipes that were broadly consistent with these recommendations in their school food guidance. Sri Lanka, The Gambia and Rwanda distinguished fresh from dried fish. Because the GDQS gram thresholds do not differentiate products by moisture content, direct comparisons with nutrient-dense dried fish portions require caution. Survey-reported fish portions in Bhutan, Costa Rica, Samoa and Timor-Leste exceeded the Intake (2025) threshold for a high contribution to meal quality.

4. Discussion

Aquatic foods illustrate the challenges of translating policy into school meals, given their diversity in nutrient content, production systems and seasonal availability, combined with differing supply chain and food safety requirements. They provide protein, omega-3 fatty acids and micronutrients important for development, with contributions depending on species and product form (FAO/WHO 2024a, 2024b; O'Connor et al., 2025). Their inclusion in school meals therefore requires translating nutritional objectives into procurement and preparation practices suited to local supply and school capacity.

4.1. Integration of Aquatic Foods into School Meals: Four Pathway Profiles

Aquatic foods were widely recognized within the documents available at each stage of the policy-to-school-plate pathway, although the availability of evidence became more limited at the operational stages. The country-level findings can be interpreted through four pathway profiles.

The first comprises ten countries with aquatic foods included in FBDGs, school food guidance, menus and recipes: Belize, Mexico, Moldova, Mongolia, Paraguay, Saint Lucia, South Africa, Sri Lanka, Timor-Leste and Vietnam. This consistency across policy and operational documents does not, by itself, verify that aquatic foods were regularly served or consumed. Among countries with corresponding survey evidence, Sri Lanka and Timor-Leste provided the clearest examples of broadly consistent aquatic food quantity across guidance, operational documents and reported provision.

The second profile comprises upstream recognition with incomplete documented operationalization: aquatic foods appeared in FBDGs or school food guidance but were absent or less consistent downstream. Albania had quantitative fish recommendations in both guidance stages but none in the reviewed menu. This may reflect implementation gaps or inaccessible current, subnational or decentralized documents.

The third profile comprises downstream inclusion without accessible national policy anchoring. The Gambia, Rwanda and Ukraine had school food guidance or operational evidence without identified FBDGs; Lao PDR included aquatic foods in recipes and reported provision despite limited upstream documentation. Local availability, procurement and program decisions may support inclusion even where national recognition is weak or undocumented.

The fourth profile comprises countries where the pathway could not be adequately assessed because the available evidence was insufficient or incomplete. No relevant policy or operational documents were identified for 31 countries, while other countries lacked one or more document types needed to assess continuity across the pathway. These findings indicate gaps in document identification or public accessibility and should not be interpreted as evidence that school food policies, school meal programs or aquatic food provision were absent. This distinction is particularly important for Pacific small island states, where fish is widely regarded as central to food security and traditional diets, although its availability and consumption vary considerably within and between countries (Bell et al., 2009; Charlton et al., 2016). For Tonga and the Marshall Islands, no relevant documents were located despite GCNF data indicating the presence of school meal programs; the Marshall Islands also contributed evidence to our stakeholder survey.

Together, these profiles distinguish documented coherence, apparent gaps between policy and operational planning, downstream inclusion without clear policy anchoring, and insufficient evidence. They extend previous work on coherence between fisheries and nutrition policies (Koehn et al., 2022) by tracing aquatic foods across dietary guidance, school food policies, menus, recipes and respondent-reported provision in multiple countries. This approach highlights both limited evidence at downstream stages and imperfect alignment between documented plans and reported provision.

4.2. Implementation Depends on Systems, Not Only Guidelines

Governance and coordination across nutrition, education, procurement, fisheries and food safety systems may influence the translation of aquatic food recommendations into school meal provision, although we did not test these relationships in this study. Survey respondents identified affordability, supply reliability, food safety, and infrastructure as factors affecting inclusion, consistent with findings from a previous scoping review of aquatic foods in school meal programs in Sub-Saharan Africa (Ahern et al., 2021). Identifying affordability as an enabler and cost as a barrier (as was the case in our stakeholder survey) may reflect different comparisons and budget constraints: aquatic foods may be affordable relative to other ASFs, while purchase, processing, transport, storage and preparation costs may still exceed school meal budgets. These findings reinforce the importance of supportive procurement systems, adequate infrastructure and budgets, and institutional capacity for implementing school food policies (Micha et al., 2018; Cohen et al., 2021; WHO 2026; Hawkins 2023). Although stronger nutrition standards can improve the nutritional quality of school meals, implementation may be constrained by cost, acceptability, availability, compliance and broader contextual factors (Cohen et al., 2021; Micha et al., 2018; WHO 2021; Ralston and Newman 2015; Vieux et al., 2013). Addressing these constraints requires a coordinated approach linking school food environments, procurement, and nutrition education (FAO 2019a).

In the surveyed settings, inclusion in planning documents did not consistently correspond with reported provision (moderate agreement, κ = 0.49), suggesting that assessment of menus and recipes alone may overestimate dietary exposure. However, the non-representative respondent-level survey sample prevents country-wide inference. Procurement, seasonal availability, budgets, supplier capacity and local decisions may help explain these discrepancies, but direct observation is needed to establish actual provision and investigate its determinants. A food systems perspective could help connect dietary objectives with feasible procurement, preparation and delivery while accounting for economic, sociocultural and environmental conditions (FAO 2024; FAO and WFP 2026).

Non-fish aquatic foods were largely invisible across the policy-to-school-plate pathway. This absence matters because “aquatic foods” encompass nutritionally and culturally diverse foods beyond finfish, including shellfish and seaweed and other algae that may be locally important sources of micronutrients, livelihoods and food system diversity (Bernhardt & O’Connor, 2021). At the same time, their composition and safety cannot be generalized: for example, seaweed nutrient content varies markedly by species and product, some products may contain excessive iodine, cadmium and inorganic arsenic, and omega-3 content may be relatively low (Aakre et al., 2021; Kendel et al., 2012; Blikra et al., 2024). Their limited visibility may reflect low consumption in some settings, aggregation within broad food groups, limited evidence, or genuine policy neglect; this review cannot distinguish among these explanations. On the other hand, their inclusion may, in some cases, represent aspirational policy framing rather than foods that are routinely available in food environments or through school procurement systems. Where locally available, culturally acceptable and safe, school food guidance and procurement assessments should consider whether non-fish aquatic foods can contribute to program objectives.

4.3. Translation Is Not Necessarily Linear, with Evidence of Both Implementation Gaps and Local Adaptation

Broad food group framing may be appropriate in an FBDG but insufficient for procurement or menu design. In this review, aquatic foods were commonly visible in FBDGs and school food guidance but grouped with other foods. Generic categories such as "protein," "animal-source foods," or "meat and alternatives" may obscure their specific nutritional contributions and offer limited incentive to prioritize them over potentially cheaper protein sources (Tlusty 2021; Ximenes et al., 2026). The appropriate degree of specificity depends on the document’s purpose. Where aquatic foods support program objectives, guidance may distinguish nutritionally or operationally important forms, such as small whole fish, oily fish or dried fish, while allowing substitution according to local availability, cost and cultural preferences. The FAO and WFP methodology for holistic nutrition guidelines and standards for school meals provides a framework for balancing specificity and flexibility (FAO and WFP 2026).

Operational guidance must also account for product form. Quantified recommendations were more visible in FBDGs and school food guidance than in menus and recipes. Frequencies generally ranged from one to three fish servings per week, while fresh fish portions frequently exceeded 40 g. Quantitative alignment was clearest in few countries: Malaysia and South Africa between policy documents; Moldova between school food guidance and menus or recipes; and Sri Lanka and Timor-Leste across guidance, operational documents and reported provision. Alignment therefore concerns both frequency and portion size; inclusion at each stage does not necessarily imply consistency in the amounts recommended or served.

School food guidance from The Gambia, Rwanda and Sri Lanka included recommendations for dried fish, with smaller portions. Drying removes water and concentrates nutrients, making dried fish more nutrient-dense per unit weight than fresh fish (Robinson et al., 2025). These portions are therefore not directly comparable with fresh fish portions or the GDQS threshold without accounting for moisture content (Intake, 2025).

4.4. Aligning Aquatic Food Products, Policy Ambitions, Procurement and School Capacity

Aquatic food guidance should reflect local food environment conditions, as product forms differ in their requirements for storage, transport, preparation, food safety, affordability, and acceptability. Integration therefore requires selecting suitable aquatic food products and, where feasible, adapting procurement systems to school capacity (Andrianarimanana et al., 2024; Ahern et al., 2025; Bonis-Profumo et al., 2024; 2025). Where cold-chain infrastructure is limited, dried, canned, smoked, or powdered products may be more feasible due to longer storage life (Kaunda et al., 2025; Mahmud et al., 2019), whereas fresh fish may suit settings with established local supply and preparation capacity. Investments in infrastructure and food safety systems may also support the procurement of other local perishable products. The balance between product adaptation and system improvements depends on local feasibility, costs, and wider benefits.

Survey respondents reported fresh products more frequently in menus or recipes (79%), followed by frozen (42%), dried (38%), canned (33%) and powdered products (17%). Categories were not mutually exclusive and indicate documented inclusion rather than serving frequency. Small fish were reported as served more often than they appeared in menus or recipes (77% versus 58%), potentially indicating provision of locally available aquatic foods not reflected in planning documents. The survey was not representative and did not capture reasons for this difference.

Procurement arrangements, food safety and cultural acceptance were recurrent considerations in survey responses. Procurement models should reflect local supply structures and institutional capacity: decentralized purchasing may strengthen links with small-scale suppliers, whereas centralized purchasing may facilitate standardized monitoring and quality control (FAO and WFP 2020). For aquatic foods, assessing fisheries value chains and supplier capacity can help identify arrangements that enable small-scale fisherfolk to supply schools (Andrianarimanana et al., 2024). Neither procurement model is inherently preferable, and hybrid arrangements may be appropriate within national programs.

Food safety requirements must protect children’s health while remaining proportionate and feasible for suppliers. Product form and recipe design can influence acceptance where bones, unfamiliar species or strong flavours are concerns (Araya and Castillo-Montes 2021; Bilbao et al., 2024; Martins et al., 2021; Nakagiri et al., 2024; Thorne-Lyman et al., 2017; Gibson et al., 2020). Fish powders and adapted recipes may improve acceptability and nutritional value in some contexts (Ahern et al., 2025; Atter et al., 2024; Donadini et al., 2013; Elolu and Ongeng 2020; Latorres et al., 2016; Laureati et al., 2016). Product selection and procurement should therefore reflect local nutritional objectives, supply conditions, safety and preferences, alongside holistic food education (Vargas et al., 2025; FAO 2020).

4.5. From Policy Recognition to Implementation: Priorities for Action and Research

Priorities depend on nutritional needs, local food systems and program capacity. Where aquatic foods are feasible and relevant, operational school food guidance can specify frequency, portions or nutritional contributions while permitting locally appropriate substitutions. Such detail may be more useful in school guidance than population-facing FBDGs.

Procurement and menu planning should reflect local production, seasonality, product form, safety, storage and supplier capacity. Cold-chain investment may be appropriate in some settings; in others, quality-assured shelf-stable products may address different constraints. Local sourcing may benefit small-scale producers, but feasibility, safety and distributional effects require context-specific assessment (Bonis-Profumo et al., 2025).

Cross-sectoral coordination may support alignment between health-led FBDGs and school guidance developed by education, health, agriculture, fisheries or social protection institutions. Centralized and decentralized models distribute procurement, preparation and management responsibilities differently; local adaptation depends on administrative, procurement and monitoring capacity (FAO and WFP 2018). This study did not classify programs sufficiently to test these relationships. Research should examine governance level, legal enforceability, coordination, monitoring and procurement.

Future research should directly assess foods offered, selected and consumed and their effects on diet quality, nutrient intake and health (FAO/WHO 2024a, 2024b; O’Connor et al., 2025). Evidence is limited in LMICs (Gelli et al., 2026), whereas studies in high-income settings have examined alignment between standards, provision and consumption (Mateo Silleras et al., 2015; Borkowski et al., 2020; Sancho Uriarte et al., 2015; Gilmour and Fairchild 2025; Juniusdottir et al., 2018). Studies should measure frequency, portions, species, product form, nutritional contribution, acceptability, plate waste, consumption and cost, using standardized tools where appropriate.

4.6. Limitations

This study has several limitations. First, the availability of policy documents, menus, and recipes varied substantially across countries, limiting comparability and restricting assessment in countries where documents were unavailable. Policy documents differed considerably in structure, terminology, and specificity. Menus and recipes may exist but are unpublished, decentralized or inaccessible. Survey responses were also based on stakeholder self-report, were not nationally representative, and likely do not reflect all schools within a country. This made it difficult to establish clear traceability between national dietary recommendations and operational meal planning.

Second, the study assessed the presence of aquatic foods within documents but not the prominence, strength or enforceability of recommendations. A passing reference to fish, a quantified consumption recommendation and a mandatory standard requiring fish twice weekly were therefore all classified as inclusion, despite implying different obligations and implementation mechanisms. Operational documents also varied substantially in scope: menus ranged from single-day examples to multi-week cycles, while recipe sources ranged from individual examples to extensive recipe books. Comparable menu-day frequencies could not be calculated because the menu documents covered different time periods and levels of completeness.

Third, we examined quantitative recommendations in relation to dietary quality thresholds (Intake, 2025), but reported provision alone does not establish whether the quantity, frequency or type of aquatic food meets nutritional objectives. Analysis of program cost structures was also outside the scope, although survey respondents frequently identified affordability as a factor affecting implementation.

Fourth, discontinuities were assessed descriptively from the availability and content of successive document types and from paired respondent reports of planned and served foods. This approach could identify apparent discontinuities but could not determine why they occurred, distinguish non-implementation from inaccessible documentation, or establish causal relationships between policy and provision.

Finally, the study did not assess implementation of other food groups or categories, so we could not determine whether the gaps identified are unique to aquatic foods or reflect broader implementation challenges. Comparative research with other nutrient-dense foods could help distinguish aquatic food barriers from wider challenges in school food systems.

5. Conclusion

Aquatic foods were widely recognized in accessible dietary and school food guidance, but downstream documents and stakeholder reports showed less consistent planning and provision. The four profiles distinguish documented continuity, weak operationalization, downstream inclusion without policy anchoring, and insufficient evidence. They guide country follow-up without equating missing documents with implementation failure; identified gaps are potential, not nationally validated.

Translation from policy to school plate requires actionable guidance, adequate resources, coordinated governance and locally appropriate procurement systems. Direct assessment of foods served and consumed—including frequency, portions, waste, nutritional contribution, cost and acceptability—is needed to establish whether aquatic foods in school meals improve children's diets and nutrition.