Assessing the implementation of public policies aimed at addressing the double burden of nutrition in Burundi: perspectives and recommendations

Abstract

Background

Non-communicable diseases (NCDs) are the leading cause of premature mortality in Africa, with Burundi facing a growing double burden of malnutrition effects: undernutrition and rising obesity rates. Despite recent policy efforts, including national strategies and increased health spending, the country’s health system remains under-resourced and vulnerable to nutrition-related health challenges.

Objective

This study aimed to assess the extent of implementation of public policies aimed at creating healthy food environments in Burundi using the Food Environment Policy Index (Food-EPI), and to identify policy gaps and priorities for addressing the double burden of malnutrition.

Methods

From December 2022 to February 2024, the Healthy Food Environment Policy Index (Food-EPI) tool was used to benchmark 56 policy and infrastructure indicators against Good Practice Statements. Documentation was reviewed, and evidence was validated by a panel of 24 experts from government, academia, and civil society through two workshops. Experts scored the level of policy implementation using a five-point Likert scale and prioritized future actions based on importance, achievability, and effectiveness.

Results

Of the 56 indicators assessed, 27% were rated as “very little or no implementation” and 37% as “low implementation.” Only 13% were rated as “high.” Infrastructure support indicators generally scored higher than policy components. Thirteen indicators were not addressed in any existing policy documents. Experts identified 13 priority actions, seven of which aligned with WHO “double-duty actions.” Inter-rater reliability was moderate (Gwet AC2 = 0.46).

Conclusions

Significant policy gaps remain in Burundi’s approach to improving food environments. There is an urgent need for comprehensive, multisectoral strategies to strengthen public health nutrition and reduce the impact of diet-related NCDs.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s40795-026-01415-z.

Introduction

Non-communicable diseases (NCDs) are the leading cause of mortality and morbidity globally, accounting for 86% of premature deaths ( between the ages of 30–70) in low and middle-income countries (LMIC) [1, 2]. Each year, more than 1.6 million people in Africa die prematurely from at least one of the four major NCDs (cardiovascular diseases, cancer, diabetes, and chronic respiratory diseases), representing 63% of NCD-related deaths on the continent [1]. Evidence has highlighted the association between the rising prevalence of nutrition-related NCDs, including type 2 diabetes and obesity, and the nutrition transition that has taken place in Africa over the past few decades [3, 4]. This transition, marked by reduced physical activity and profound changes in the food environments that favor the consumption of unhealthy, highly processed foods and beverages, has been increasingly associated to a higher risk of NCDs [5–7].

Food environments are defined as the collective physical, economic, policy and sociocultural conditions that influence food and beverage choices and ultimately affect nutritional status [8]. For instance, modern retailers – such as supermarkets - typically offer a higher proportion of processed foods than traditional markets [9]. Today, evidence highlights the important role of effective public policies in creating healthy food environments to reduce the high burden of malnutrition in all its forms - both undernutrition (i.e., micronutrient deficiencies, underweight, and childhood stunting and wasting) and overnutrition (i.e., overweight and obesity) [5, 8, 10–12]. In response to the rising of NCDs, major international intergovernmental organizations, including the World Health Organization (WHO), are calling on countries to adopt a system-level approach that extend beyond individual behavior change to include fiscal, agricultural, trade, and food marketing policies [13]. To support these efforts, the WHO has issued multiple policy frameworks and guidance documents, such as the Global Action Plan for the Prevention and Control of NCDs 2013–2030, the “Best buys” to address NCDs1, the Acceleration Plan to Stop Obesity, and guidance on “Double-duty actions” to address multiple forms of malnutrition simultaneously [10, 14–16].

Evidence from country experiences demonstrates the potential effectiveness of such policies. In Chile, for example, the implementation of the Food Labelling and Advertising Law was associated with a substantial reduction in the proportion of food and beverage products high in sugar and sodium [17]. While the evidence on the effectiveness and cost-effectiveness of these policy interventions continues to grow [10, 17–20], countries worldwide still face challenges in translating this evidence into action to improve food environments, dietary patterns and population health [21]. One major challenge is interference from the food and beverage industry, which uses a range of strategies to influence nutrition policies [22]. These include shaping the evidence base through industry‑funded research, promoting deregulation by emphasizing scientific uncertainty, and shifting responsibility for diet‑related harms onto individuals, parents, and physical inactivity. Industry actors also consistently highlight their economic importance and promote voluntary and self‑regulatory approaches as alternatives to binding public health policies [22, 23].

As a result, policy implementation has been slow and insufficient, with a focus on behavioral change rather than addressing the underlying food environment determinants of unhealthy diets [24, 25]. Moreover, policies have not been properly evaluated to ensure their effectiveness [26].

Like many African countries, Burundi is facing a double burden of malnutrition, characterized by the co-existence of undernutrition and overnutrition and diet-related NCDs [4, 5, 27–29]. Between 2000 and 2016, NCDs accounted for 32% of all deaths in Burundi; and in 2021, the probability of premature mortality from an NCD was 25% [1, 30]. Among adults, an estimated 11% of women and 3% of men are affected by obesity, with evidence indicating that the prevalence is on the rise [27, 31]. At the same time, undernutrition remains a major public health concern, with 51% of children under five years of age affected by stunting and 6% by wasting [27].

Over the past two decades, Burundi has implemented policy measures to prevent and control NCDs and improve nutrition nationwide. NCDs are identified as a priority health objective in both the National Development Plan 2018–2027 and the National Health Policy 2021–2027 [32, 33]. In addition, the National Program for the Control of Chronic Non-Communicable Diseases was established in 2009 to strengthen and coordinate country-level actions [34]. This growing policy attention to NCDs has coincided with increased health spending over the last decade, although spending remains below the Abuja Declaration target of allocating 15% of the national annual budget to the health sector [35].

Despite these efforts, Burundi’s health system is largely underfunded and highly dependent on external aid [36]. An estimated 30% of healthcare expenditure comes from household out-of-pocket payments, which exacerbates socioeconomic inequities and limit access to care [36, 37]. Persistent resource constraints further weaken the health system, which is not yet adequately equipped to cope with the rising incidence of NCDs [31]. Taken together, these conditions underscore both the urgency of addressing malnutrition and diet-related NCDs and the opportunity for the Government of Burundi to strengthen and scale up preventive policies.

A starting point for policy change is the systematic assessment of existing policies that influence food environments, with the aim of identifying gaps and weaknesses relative to international best practices. The Healthy Food Environment Policy Index (Food-EPI) is a useful tool to monitor government action, identify policy inconsistencies, and create a system of accountability for improving healthy food environments [38]. It also raises awareness of the ways in which food environments influence diets [8, 38]. Over the past five years, the Food-EPI methodology has been increasingly used in sub-Saharan Africa [39–42].

The objective of this study is to assess the extent to which public policies aimed at creating healthy food environments have been implemented in Burundi, to identify key policy gaps, and to formulate recommendations to support government action in addressing the country’s double burden of malnutrition.

Methods

Food-EPI overall methodology

This study was conducted from December 2022–February 2024 using the Food-EPI framework and methodology developed by the International Network for Food and Obesity/Non-communicable Diseases Research, Monitoring and Action Support (INFORMAS). The framework was designed to assess the extent of implementation of government policies and actions aimed at creating healthy food environments to reduce obesity, diet-related NCDs, and their related inequalities [8]. Its components are grouped into two categories - “Policy” and “Infrastructure Support” - and 13 domains (Fig. 1 - Healthy Food Environment Policy Index). Food-EPI provides a systematic process for collecting, validating and analyzing public policies that influence food environments and dietary patterns. These policies are assessed against a set of indicators derived from international best practices. By comparing these policies with the benchmarks, Food-EPI facilitates the identification and prioritization of policy actions needed to strengthen food environments and support healthier diet [8].

Fig. 1.

Domains covered by the 56 Epi-Food indicators included in the study. This figure was extracted and adapted from Swinburn et al. [8]. https://doi.org/10.1111/obr.12073

In accordance with the Food-EPI methodological guidelines, the study was conducted in eight consecutive phases (8): (1) analyzing the context and adapting the Food-EPI tool to the Burundi context; (2) collecting documentation on policies aimed at creating healthy food environments for the prevention of malnutrition; (3) drafting an evidence report on the extent of implementation of food environments policies; (4) validating the evidence report by an expert panel; (5) expert panel rating of the level of implementation of government policies; (6) weighting and calculating Food-EPI scores; (7) identification and prioritization of actions by the expert panel for recommendation to the government and; (8) analyzing the data and synthesizing the main results.

Step 1: adaptation of the food-EPI methodology

We made three changes to the original Food-EPI methodology to better reflect the Burundian context. First, while the Food-EPI framework was originally designed to assess public policies related to obesity, a growing number of assessments in Africa have expanded the framework to include 12 indicators related to undernutrition, bringing the total number of indicators to 59 [39, 40, 42–44]. This adaptation enhances the framework’s relevance for examining the double burden of malnutrition prevailing in LMICs [40, 42, 43]. Our study included 56 indicators: 47 from the original Food-EPI tool and nine related to undernutrition (Appendix 1). First, to simplify the wording and reduce perceived redundancy, an expert panel (see below in Constitution of the Expert panel) recommended a set of revisions. As a result, five undernutrition indicators within the “Leadership” domain were modified: (i) three indicators addressing breastfeeding and complementary feeding were merged into a single indicator; and (ii) two indicators related to government leadership in support of healthier food environments were combined into one (Appendix 1). Second, experts were asked to compare evidence collected on “Policies” and “Infrastructure Supports” against the corresponding Food-EPI Good Practice Statements (as opposed to International best practices). These Good Practice Statements represent ideal actions by governments to create healthier food environments [45]. Third, we organized a two-day in-person workshop with the Expert panel to facilitate the validation of evidence on public policies addressing the double burden of nutrition, learning from experiences from Benin and Burkina Faso [40, 43].

Step 2 and 3: documentation and evidence report

We conducted a review of relevant government policies, strategies and plans related to nutrition and health that may influence nutritional outcomes, including government budgets, legislation, and trade regulations. We gathered information from the official websites of the Burundi government, local universities, and international organizations (the WHO, the Food and Agriculture Organization, the World Bank, etc.). Additional documentation was obtained from government officials, university professors and civil society organization representatives. For each of the 56 indicators, evidence of implementation (or lack thereof) was collected, reviewed and documented in a draft evidence report.

Step 4: evidence validation

Constitution of the expert panel

Through Google searches, recommendations from the Ministry of Health and the research team’s network, we identified 43 experts in nutrition, public health, food systems and other nutrition-sensitive domains. An invitation letter and the project description were sent to them the week of 29 January 2024, and follow-up messages and calls were initiated a week later.

Evidence document validation

An in-person validation workshop was held on 13–14 February 2024 in Bujumbura, Burundi to systematically review and validate the collected evidence. Of the 43 experts identified, 20 were invited to participate in the workshop. Participants were purposively selected to ensure balanced representation across key stakeholder groups while maintaining a manageable group size to facilitate meaningful interaction and discussion. The workshop began with an orientation session led by a member of the research team, during which the project objectives, scope and the Food-EPI methodology were presented. Subsequently, each of the 56 Food-EPI indicators was reviewed in plenary, together with the corresponding Good Practice Statements and the summarized evidence (or lack thereof) documented in the draft evidence report. The expert group identified missing and no longer relevant documents, suggested revisions, and proposed additional experts to consult. Following the validation workshop, feedback from the experts was incorporated, and a final evidence report was produced.

Step 5: rating the extent of policy implementation

We convened the expert panel to assess the extent of policy implementation relative to the Food-EPI Good Practice Statements during an in-person workshop held on 27 February 2024 in Bujumbura. Invitation letters were sent to the 43 experts identified during the previous phase of the study. Prior to the workshop, the revised evidence report was shared by email to all invited experts, who were asked to read it in advance of the meeting.

Two members of the research team facilitated the workshop. First, they presented the project objectives and methodology, and an illustrative example of the rating process to familiarize the participants with the exercise. Participants were then randomly assigned to one of two working groups corresponding to the two main components of the Food-EPI framework: One group worked on the Policy component, while the other focused on the Infrastructure Support component. Each group was supported by a member of the research team, whose role was to provide methodological clarification and ensure adherence to the Food-EPI protocol without influencing participants’ assessments.

Within their respective groups, participants systematically reviewed and discussed the Food-EPI indicators, the associated Good Practice Statements, and the supporting evidence documented in the evidence report. Following a group discussion on each indicator, participants independently rated the level of implementation using the Food‑EPI evaluation form. Participants were asked to individually rate the level of implementation as compared to the corresponding Good Practice statement, taking into account 2 additional criteria: (1) the quality of government policies and actions compared to the Good Practice Statements; and (2) the extent of implementation of all aspects of the policy cycle, including the government’s intentions, plans for implementation, and actions that were already underway.

A five-point Likert scale was used, with scores ranging from 1 (less than 20% implementation) to 5 (80–100% implementation). A score of 6 was assigned in cases where participants considered an indicator “impossible to rate.”

Step 6: analysis

Descriptive analyses of the expert panel characteristics (age, sex, professional background, and years of experience) were conducted using Excel. In accordance with the Food-EPI protocol, indicator ratings were grouped into four implementation categories: High (> 75%), Medium (50–75%), Low ( 25–50%) and very little or no implementation (≤ 25%). These categories were used to determine the overall level of policy implementation across domains. Inter-rater reliability was assessed using the Gwet AC2 coefficient, calculated in Excel, to evaluate the level of agreement between the raters.

Step 7 and 8: identification and prioritization of actions

Following the policy implementation assessment, the expert panel identified and prioritized actions to recommend to the government to strengthen food environments. This activity was conducted during a workshop on 28 February 2024. The same members of the research teams moderated the session to ensure methodological consistency. They presented the preliminary results from the previous day assessments and the Food‑EPI prioritization process. Using the Food-EPI prioritization form, the experts individually formulated concrete actions, taking into account three criteria: importance, achievability, and perceived likelihood that the action would reduce the double burden of malnutrition. Priority actions were then agreed upon within each group and subsequently presented and discussed in plenary.

Ethics

This study was approved by the Science and Health Research Ethics Committee of the University of Montreal on 19 December 2022 under the project number 2022–3909. All study participants provided written consent before taking part in the project. Participants received a small token of appreciation, valued at 46 CAD per day, for their time and contribution to the project.

Results

Policy scan and evidence

A total of 75 documents were identified and screened for stated objectives, targets, and references to one or more of the 56 indicators under review. Of these, 56 documents were judged to be relevant and were included in the evidence report (Appendix 2). The review assessed the extent to which the government has developed or is in the process of developing policies to improve population nutrition, using the Food-EPI Good Practice Statements as benchmarks. Overall, we found 13 indicators (11 in the Policy domain and 2 in Infrastructure Support) that were not explicitly mentioned or addressed in any of the policy documents reviewed (Table 1). These included indicators on labelling on food packaging and menu boards in restaurants, food promotion, food price, food provision, retail, governance and health in All Policies.

Table 1.

Food-EPI indicators for which no policy guidance was found

Policy domain
	Food labelling	Menu labelling
	Food promotion	Promotion to children in non-broadcast media
		Promotion to children in children’s settings
	Food price	Food-related income support
	Food provision	Promote Healthy foods in public workplaces
		Promote Healthy foods in private workplaces
	Retail	Planning policies for unhealthy food outlets
		Planning policies for healthy food outlets
		Food availability in stores
		Food availability and promotion in restaurants
		Policies on informal retailers
Infrastructure domain
	Governance	Transparency in policy development
	Health in all policies	Health impact assessments in non-food policies

Characteristics of the expert panel

A total of 43 experts were invited to participate in the Food-EPI process. Of these, 12 did not respond to the invitation and seven declined. In total, 24 experts attended at least one of two 2-day workshops. Thirteen experts participated in the evidence validation workshop, while 21 took place in the rating and prioritization workshop. Ten participants attended both workshops. Participants included representatives from government ministries agencies (n = 10), civil society organizations (n = 8) and academia (n = 6) (Table 2). Out of these 24 experts, 19 participated in the rating exercise.

Table 2.

Characteristics of experts who participated in the workshops

Characteristic		Number	Proportion
Affiliation
	Government	10	42%
	Academia	8	33%
	NGO	6	25%
Age
	< 30 years	2	8%
	30–45 years	7	29%
	46–55 years	6	25%
	> 55 years	9	38%
Experience
	< 10 years	2	8%
	10–20 years	12	50%
	> 20 years	10	42%
Sex
	Female	6	25%
	Male	18	75%
Total		24	100%

Extent of implementation and priorities

Among the 56 indicators, 15 (27%) were rated as having “very little or no implementation,” while 21 (37%) received a “low implementation” rating. Thirteen indicators (23%) were assessed as having “medium implementation,” and only seven (13%) were rated “high implementation” (Fig. 2A–B).

Fig. 2.

A Average score for each indicator’s implementation level (policy). The x axis represents the average score for each indicator’s implementation level, as rated by the 19 participants on a scale of 0 to 5. The average score is therefore a continuous numerical value between 0 (minimum) and 5 (maximum). These average scores were categorized and put in colors according to the scale commonly used by FOOD-EPI: 0–2 none/very low ; 2–3 low ; 3–4 medium ; 4–5 high. * indicates an indicator for which no mention of implementation was found during the documentation review. B Average score for each indicator’s implementation level (infrastructure). The x axis represents the average score for each indicator's implementation level, as rated by the 19 participants on a scale of 0 to 5. The average score is therefore a continuous numerical value between 0 (minimum) and 5 (maximum). These average scores were categorized and put in colors according to the scale commonly used by FOOD-EPI: 0–2 none/very low ; 2–3 low ; 3–4 medium ; 4–5 high. * indicates an indicator for which no mention of implementation was found during the documentation review

Within the Policy domain, 22 out of 28 indicators (79%), were rated “very little or low implementation.” Indicators in the “very low or no implementation” category (n = 13) include Front-of-package labeling; menu labelling in restaurant settings; restrictions on the promotion of unhealthy food in broadcast media; increased taxation on unhealthy food; food-related income support; zoning regulations for healthy and unhealthy food outlets; and policies addressing informal food retailers. In addition, nine indicators were rated as having “low implementation”, and five received a “medium” rating. These primarily addressed regulation of infant formula marketing, reduced taxation on healthy foods, school nutrition policies for healthy food provision, policies to improve water, sanitation, and hygiene as well as measures to protect regulatory capacity for nutrition. The impact of trade agreement was the only Policy indicator rated as having “high implementation”.

In contrast, the indicators within the Infrastructure Support component had higher levels of implementation overall. Six indicators were rated as having “high implementation”: (1) government leadership to improve food environments; (2) leadership in the promotion of exclusive breastfeeding; (3) use of evidence in food policy development; (4) evaluation of major nutrition programs and policies; (5) monitoring of breastfeeding indicators; and (6) government coordination to support policy coherence. Eight Infrastructure indicators received a “medium implementation” rating, while 12 fall into the category of “low implementation.” Two indicators - transparency in policy development and assessment of health impacts in non-food policies - were rated “very little or no implementation.”

The inter-rater reliability coefficient was 0.46 (95% CI: 0.34–0.59), indicating moderate agreement among raters.

Finally, the experts identified 34 potential actions to improve food environments (n = 28) and reduce undernutrition (n = 6). Of these, eight actions within the Policy Component and five in the Infrastructure Component were prioritized as having both “higher importance” and “higher achievability” (Table 3). Notably, family planning was identified as a priority action.

Table 3.

Priority actions recommended by the expert panel, by component^¶

Domain	Priority recommendation
Policy Component
1. Food Composition	Build capacities to strengthen food inspection and certification services
2. Food Labelling	Enhance labelling on food packaging and menu boards in restaurants and food traceability systems
3. Food promotion	Restrict promotion of unhealthy foods to children (< 16 years) across all media
4. Food Prices	Subsidise the production and sale of healthy foods
5. Food Provision	Scale up school canteens and diversify the meals served
6. Food retail	Implement regulation on the density of fast-food outlets (zoning)
7. Other	Enhance Family planning programmes
8. Other	Strengthen nutrition rehabilitation center
Infrastructure Component
9. Leadership	Strengthen multi-sectoral collaboration to tackle all forms of malnutrition for the entire population
10. Leadership	Develop and disseminate a comprehensive food guide for adults, children and infants
11. Leadership	Strengthen poverty reduction and malnutrition programmes including income-generating activities for vulnerable populations
12. Monitoring	Set up community-based children development centres
13. Funding	Increase government funding for nutrition and NCD prevention

^¶Experts were asked to base their recommendations on 3 key dimensions: importance, achievability and potential effect

Evaluation of the food-EPI tool

Sixteen experts completed the Food-EPI post-workshop evaluation questionnaire. Of these, 56% identified as government officials, while the remaining 44% identified as researchers or civil society representatives. When asked whether the project would contribute to policy change in Burundi, 94% “agreed” or “strongly agreed”. All 16 (100%) participants indicated that the Food-EPI process had improved their understanding of the food environments and related food and nutrition policies. Participants also reported increased awareness of current good practices related to food environments. In addition, 44% of the experts found it “somewhat difficult” to assess the extent of policies implementation. When asked about the comprehensiveness of the Food-EPI tool, 37% of respondents felt that the tool was “not comprehensive enough” or were “neutral.”

Discussion

The prevalence of overweight and obesity is on the rise in Burundi, alongside high levels of undernutrition, particularly for children under five years of age [27]. Using the Food-EPI tool, we benchmarked and assessed the extent of which nutrition-related policies are implemented against internationally recognised Good Practice Statements. Our study is among the few in Africa to apply the Food-EPI framework to evaluate policies addressing the double burden of malnutrition [39, 40, 42, 43]. To our knowledge, three Food-EPI studies (in Benin, Burkina Faso, and South Africa) have benchmarked public policies against all nine undernutrition indicators included in our study [40, 42].

Our findings suggest that, while efforts to improve nutrition in Burundi are increasing, significant implementation gaps remain. Notably, policies aimed at preventing diet-related NCDs are beginning to emerge. Established in 2009, the National Program for the Control of Non-communicable Diseases coordinates government action on NCD prevention, including policy development and implementation. Similar increases in attention to NCD prevention were observed across several African countries in the late 2000s, although widespread gaps in policy adoption and implementation persist [46].

Overall, 64% of the assessed policies were rated as having “low” implementation or “nonexistent” when compared with the Food-EPI Good Practice Statements. Comparative findings have been reported in South Africa (70%), Senegal (72%) and Ghana (75%) [39, 42, 44]. Benin and Kenya findings show higher percentages, where 86% and 84% of indicators, respectively, fell into this category [40, 47]. In contrast, the Burkina Faso study reported a lower rate at 46% [43].

In our study, 23% (13 out of 56) of the indicators were rated as having “medium” implementation. This result is similar to findings from South Africa (22%) [42]. In addition, seven (12%) indicators were rated as having “high implementation”, including one from the Policy component and six within the Infrastructure Support component. This aligns with results from South Africa (8%) and Burkina Faso (10.7%) [42, 43]. By comparison, Food-EPI studies in Benin, Kenya and Senegal reported no indicators with “high implementation”, while Ghana reported only one, related to government action to restrict the marketing of infant formula [39, 40, 44, 47].

The relatively high number of indicators rated as “high implementation” in Burundi may be explained by the composition of the expert panel. For example, assessment of trade agreement impacts was rated as “high implementation”. This may reflect the fact that 42% of reviewers were government officials, who may have assessed the policies more favourably. This was also observed in other Food-EPI studies that included both government officials and independent actors in the rating process [40, 43, 48, 49].

Overall, findings related to undernutrition indicators (n = 9) are consistent with results from other LMIC countries [40, 42, 43]. Of these indicators, the highest percentage of implementation were “breastfeeding and complementary feeding targets” and “monitoring of breastfeeding and complimentary feeding targets”. This is likely attributable to decades of sustained global efforts to improve maternal and child health through the promotion of breastfeeding. For example, in line with the International Code of Marketing of Breast-milk Substitutes, the WHO has encouraged Member States to establish effective, independent, and transparent monitoring systems to support breastfeeding and enforce the Code’s provisions [50]. The level of implementation was lowest for four indicators: “Food fortification with micronutrients”, “policies promoting breastfeeding in the workplace and in public spaces”; “policies on informal retailers” and “food safety indicators”. Tapsoba et al. reported four undernutrition indicators in the “high implementation” category of which two are the same as the ones in Burundi [43]. Similarly, policies to improve water and sanitation programmes were rated “medium” in both Burundi and Benin [40]. Furthermore, “Food fortification programs” and “food safety” indicators scored “low” for the two countries [40].

While the Food-EPI process in Burundi revealed policy actions aimed at preventing all forms of malnutrition and diet-related diseases, it also highlighted policy gaps in 13 Food-EPI indicators (11 in the ‘Policy’ component and two in the ‘Infrastructure Support’ component). This indicates opportunities for the Government of Burundi to strengthen its nutrition policy actions by implementing proven interventions such as the WHO’s “best buys”, the Acceleration Plan to Stop Obesity, and front-of-package labelling guidelines. Except for policies promoting breastfeeding practices, the majority of the WHO “best buy” actions targeting unhealthy diets appear to be lacking. These include fiscal policies to support healthy diets, front-of- package labelling, policies on public food procurement, and policies to protect children from the harmful impact of food marketing. Food-EPI studies in Benin, Ghana, Senegal, Bangladesh, Pakistan and Sri Lanka found similar results [39, 40, 44, 51].

Of the 13 priority actions recommended by the Expert Panel, one falls outside the Food-EPI indicators: enhancing family planning programs. Burundi’s high population density (which increased by 283% between 1961 and 2017) and continued population growth (estimated at 2.3% annually) put pressure on arable land, which affects diet and nutrition outcome [29, 52]. Other Food-EPI studies have reported recommendations on social determinants of health. For instance, recommendations to the Canadian government included improving access to affordable housing and providing basic income for all people living in Canada [45]. Also of note is that six (46%) of the priority actions focused on improving undernutrition, including strengthening nutrition rehabilitation centers for children under five years. This likely reflects the persistence of high levels of undernutrition in Burundi, particularly in rural areas and regions with high population density, as well as the limited affordability of healthy foods for many families [53, 54]. Another possible interpretation is that, like other LMICs, the Burundian government has had a stronger mandate to address undernutrition. This emphasis is reflected in strategies and plans to a greater extent than overnutrition [55]. The findings also indicate that seven out of the 13 recommended actions are aligned with four of the five WHO’s double-duty actions [16]: interventions to optimize early nutrition, enhance maternal nutrition and antenatal care programs, strengthen school food policies and programs, and regulate food marketing. Consistent with findings from other assessments in Africa, there is an opportunity to further expand and use the Food-EPI methodology in settings facing the double-burden of malnutrition [39, 40, 42–44, 47].

Lastly, nineteen of the 43 invited experts (44%) participated in the policy implementation rating exercise. This participation rate is comparable to those reported in Food-EPI studies in Guatemala (45.1%), Singapore (45.1%), and Chile (45.5%) [11]. Our study also suggests that the Food-EPI process helped increase capacity among participants. Sixteen participants (84%) completed the Food-EPI evaluation questionnaire, and 62% of them agreed that their knowledge of food environments and related food and nutrition policies had increased as a result of the project. Furthermore, 44% of respondents found scoring the level of policy implementation to be somewhat difficult. This may be explained by the fact that the concept of food environments was new to many of them, or that their expertise was aligned with only a subset of Food-EPI indicators. In addition, some respondents perceived the scoring process as subjective, particularly where discrepancies existed between policies on paper and their actual implementation. Overall, while the Food-EPI process appears to have increased awareness and understanding of food environments, the findings also point to important capacity gaps and opportunities to advance work in Burundi to support healthier food environments.

Strengths and limitations

We took steps to ensure the credibility and rigor of this study; however, some limitations remain. Relevant policy documents may have been missed despite efforts to identify all materials. In addition, the concept of healthy food environments was new to some participants, which may have influenced their assessment of policy implementation. A key strength of this study is its capacity-building effect: 62% of experts who evaluated the Food-EPI process reported improved knowledge of healthy food environments. The study also incorporated undernutrition indicators, allowing for a more comprehensive assessment of policies addressing the double burden of malnutrition.

Conclusion

This study is the first to assess the food environments in Burundi using the Food-EPI tool, providing an overview of existing policies, plans and programs aimed at addressing malnutrition in all its forms. The findings highlight important opportunities to strengthen the food environments in order to deliver nutritious, affordable and culturally appropriate foods. This can be achieved through the implementation of policies that simultaneously address undernutrition and diet-related NCDs, alongside reinforcing the infrastructure support environment required to effectively impement these actions. The persistent rates of undernutrition and the rising burden of diet-related NCDs in Burundi, coupled with the underperforming health system call for urgent governmental actions to improve public policies and strategies to combat the NCD epidemic.

Authors' contributions

FS, JSM, JCM, and TD designed the study. FS performed the documentation screening and gathering. FS and TD organised and oversaw the expert panels and the priority exercise. FS and TD analyzed the data. FS prepared the first draft of the manuscript. All authors read, revised and approved the final draft.

Funding

This project received funding from the Réseau de recherche en santé des populations du Québec in 2023. The funding agency had no role in the design of the study; the collection, analysis, and interpretation of data; in the writing of the draft; or in the decision to submit the manuscript.

Data availability

All documents screened and reviewed during this study are publicly available.

Declarations

Ethics approval and consent to participate

This study was approved by the Science and Health Research Ethics Committee of the University of Montreal on 19 December 2022, under project number 2022–3909. All study participants provided written consent before taking part in the project.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.