Interventions and policies aimed at improving nutrition in Small Island Developing States: a rapid review

ABSTRACT

Objective.

To describe features of nutritional interventions implemented in Small Island Developing States (SIDS) in the past 20 years.

Methods.

A rapid scoping review was conducted by searching PubMed and Web of Science databases for interventions conducted in SIDS that sought to improve the nutrition of their populations between 2000 and 2019 inclusive. The Noncommunicable diseases progress monitor 2020 was also examined to assess nutritional policies in SIDS.

Results.

A total of 174 interventions were implemented in 49 of the 58 SIDS. The greatest proportion were conducted in the Caribbean (75 interventions; 43%), with the Pacific region, and the Atlantic, Indian Ocean, Mediterranean and South China Sea region each implementing about 30% of interventions. Using the NOURISHING framework, most interventions (67%) were implemented at the community and national or policy level, using multiple components of the framework. The greatest proportion of interventions (35%) were educational and awareness raising. Restrictions on physical availability of and increased taxation on alcohol were the most commonly reported policies that were partially or fully achieved; restrictions on fats were the least commonly reported. These findings were generally consistent across the SIDS regions.

Conclusions.

There is a paucity of nutritional policies in SIDS; the reasons may be linked to their social, economic, and environmental vulnerabilities. Interventions should be expanded beyond education to encompass multiple components of the NOURISHING framework, with multisectoral inclusion to ensure stronger stakeholder collaboration and buy-in. A systematic review is warranted using a fuller range of sources to assess the effectiveness of interventions.

Small Island Developing States (SIDS) is a term designated by the United Nations that refers to a group of 58 small island countries located in the Caribbean, the Pacific, and the Atlantic, Indian Ocean, Mediterranean and South China Sea (AIMS), which share complex social, economic, and environmental vulnerabilities (1, 2). These vulnerabilities contribute partly to their high burden of chronic noncommunicable diseases (3, 4). For instance, the prevalence of child stunting in five of the poorest SIDS exceeds 20% (4). On the other hand, more than half of SIDS report a prevalence of adolescent obesity of ≥10% and adult obesity of ≥20%, with the highest rates found in the Pacific region (5). For example, 32% and 61%, respectively, of adolescents and adults in Nauru are obese (5). In addition, Trinidad and Tobago has the highest prevalence of diabetes in the Caribbean, with 46% of adults suffering from the condition (5).

Amidst an environment of unstable food supply and accessibility, food utilization in SIDS is characterized by nutritionally poor food choices (4). A nutrition transition has occurred where locally grown foods have been largely replaced by imported foods, which include many calorie-dense, processed items (6, 7). Yet, consumption of good-quality local food is critical, not only for food security and population health, but also for national economic development, as it is a means of generating income through local agriculture and, at the same time, reducing the economic burden of diet-related noncommunicable diseases (6).

The 2017 Global Action Programme on Food Security and Nutrition in SIDS has a three-pronged approach to improving food and nutrition security, namely: strengthening the enabling environments for food security and nutrition; improving sustainability, resilience, and nutrition-sensitivity of food systems; and empowering people and communities for food security and nutrition (4). Food labeling, nutrition standards and guidelines, food marketing, nutritional quality of the food supply, trade and fiscal policies, and food chain incentives are all potential areas for action (8). This interdisciplinary approach is argued to be especially necessary in the field of nutrition as it is related to economic, political, and environmental sciences and it influences humans at physiological, social, and environmental levels (9).

With this backdrop of a high noncommunicable disease burden and poor food utilization, and well documented approaches to improve nutrition in SIDS, it is important to understand what is being done on the ground within these countries to reduce this malnutrition burden. As no reviews have been published that examine what is being done in SIDS to improve the population-level nutritional status, this rapid review sought to document, classify, and describe the features of nutritional interventions that have been conducted in SIDS in the past 20 years, from the individual level to the policy level. The results of the review are being used to inform the Food and Nutrition project (10) on developing its nutrition interventions. This 4-year project focuses on investigating and influencing national food systems to help combat obesity and diet-related noncommunicable diseases in the Caribbean Community (CARICOM), through stakeholder interaction, extensive research, and the creation of several wide-reaching interventions.

METHODS

Study design

This scoping review was expedited as a rapid review by using fewer databases and not searching the grey literature. It followed the guidance of Cochrane training on rapid reviews, the PRISMA extension for scoping reviews, and a leading expert in scoping reviews (11–13).

Eligibility

We searched for reports that described a nutritional intervention conducted in SIDS between 2000 and 2019. These interventions could be at the individual, household, community, or policy level. Study design or study outcomes were not used as eligibility criteria. Inclusion criteria were interventions that: were conducted in any of the 58 SIDS, as designated by the UN (1); sought to improve nutrition of human populations; included a study population of any age; were published between 2000 and 2019, inclusive, or were unpublished; and were published in English.

Search strategy

MEDLINE (via PubMed) (14) and Web of Science (via Clarivate) (15) databases were searched, and the reference lists of included records were also screened for any potentially relevant reports. Search terms were formulated through thorough research of the topic and subsequent rounds of drafting and editing by academic researchers with experience in nutrition and food security, and systematic reviews. Search strategies were kept broad to be highly sensitive and are listed in Annex 1 of the supplementary material.

Record selection

Record selection was done in two stages. First, titles and abstracts were screened for potentially relevant (include) or non-relevant (exclude) records. Second, the full-text records included after the initial screening were reviewed to determine if they were relevant (include) or not relevant (exclude). The records were reviewed by two independent reviewers to determine eligibility. Any disagreements or uncertainties about records at the title/abstract screening stage were resolved by automatically including these papers. During full-text screening, any disagreements were resolved by discussion between the researchers. Records were managed in Rayyan reference manager (16).

Data extraction

A data abstraction form was developed in REDCap© 7.3.4 (17), guided by the CONSORT statement on transparent reporting of trials (18). Data were extracted by two independent reviewers. Publication details (author, year, title, and journal), country, sample size and type, and level and type of intervention were extracted during the secondary screening, with instances of discrepancy resolved by discussion between researchers. Classification of systems-level interventions followed the NOURISHING framework, which identifies 10 areas across three domains (food environment, food system, and behavior change communication) where governments can influence what their populations eat (19). Classification of individual-level interventions was done on a case-by-case basis.

Synthesis of results

The review takes the form of a narrative synthesis of evidence. Key record-level information was summarized for all included records. To supplement findings from the included records, the WHO Noncommunicable diseases progress monitor 2020 was used to identify nutrition policies that have been adopted in SIDS (20).

RESULTS

Summary of included records

Of 13 206 records identified through database searching, 158 were included after full-text screening; five were also identified from the reference lists (Figure 1). The full list of the 163 records can be found in Annex 2 of the supplementary material. Within these 163 records, 174 unique interventions were found. Records that discussed the same intervention were grouped as one for the analysis.

FIGURE 1. Flow diagram of search strategy and record selection.

Source: prepared by the authors based on PRISMA recommendations (13).

Interventions were conducted in 49 of the 58 SIDS. Overall, five of nine (56%) SIDS in the AIMS region, 24 of 29 (83%) SIDS in the Caribbean region and 20 of 20 (100%) SIDS in the Pacific region conducted interventions. The greatest proportion of interventions were conducted in the Caribbean (75 (43%)), with the Pacific region and the AIMS region each hosting about 30% of the interventions (Table 1A, supplementary material). Countries that had the most interventions were Fiji (26 interventions), and Guinea-Bissau, Singapore and Jamaica (each with 21 interventions). Participants were most frequently recruited from the general population (46% of interventions), followed by health clinics (34%) and schools (15%); workplaces accounted for 5% of interventions.

Types of interventions

Of the 174 unique interventions, 57 (33%) were individual or household-level interventions, 45 (26%) were community-level interventions, and 72 (41%) were national or policy-level interventions. Classification of the 117 community-level and national or policy-level interventions according to the NOURISHING framework is shown in Table 1.

TABLE 1. Community-level and national- or policy-level interventions, classified by region of Small Island Developing States and components of the NOURISHING framework (19).

Domain	Intervention component	Number of interventions			Total no. (%) (n = 200)
		AIMS	Caribbean	Pacific
Food environment	N1 (Nutrition label standards and regulations on the use of claims and implied claims on food)	2	5	2	8 (4)
	O (Offer healthy food and set standards in public institutions and other specific settings)	5	8	11	23 (12)
	U (Use economic tools to address food affordability and purchase incentives)	1	7	12	19 (10)
	R (Restrict food advertising and other forms of commercial promotion)	2	1	5	8 (4)
	I1 (Improve nutritional quality of the whole food supply)	3	7	12	19 (10)
	S (Set incentives and rules to create a healthy retail and food service environment)	2	2	8	12 (6)
Food system	H (Harness supply chain and actions across sectors to ensure coherence with health)	2	9	9	18 (9)
Behavior change communication	I2 (Inform people about food and nutrition through public awareness)	8	13	15	33 (17)
	N2 (Nutrition advice and counselling in healthcare settings)	2	11	3	17 (9)
	G (Give nutrition education and skills)	9	15	14	37 (19)
	Other	0	4	2	6 (3)

AIMS, Atlantic, Indian Ocean, Mediterranean and South China Sea.

Notes: Interventions could be classified by more than one NOURISHING component or may occur in multiple countries of different regions. Shading of cells indicates areas of high (darker grey) versus low (lighter grey) numbers of interventions.

Source: prepared by the authors from the results of the review.

The most commonly reported intervention components were giving nutrition education and skills (G), followed by raising public awareness about food and nutrition (I²). The least common intervention components were setting nutrition label standards and regulations on claims about food (N¹), and restricting food advertising and commercial promotion (R). The distribution of intervention components was relatively consistent across SIDS regions, except the use of economic tools (U), which was less common in the AIMS region.

Interventions at the community or national/policy level tended to have multiple NOURISHING components (up to eight). Interventions with five or more components were mostly policy-driven and related to the intake of salt, fat, and fruits and vegetables. For example, Fiji implemented a 4-year multisectoral intervention through health workers, media, community leaders, food manufacturers and retailers, and consumers to reduce national salt consumption (21). In Mauritius, efforts were made to reduce fat consumption through food labelling, promotion of dietary guidelines, curriculum-based education in schools, and regulation of fatty food provision from government-run food services and at government functions (22). The Marshall Islands implemented the Healthy Stores Program which targeted food stores to encourage healthier food choices and cooking methods through: tastings and cooking demonstrations; labeling and restocking of shelves with healthier food items; and in-store and out-of-store media messaging (23). Table 2 lists examples of interventions classified by the NOURISHING framework.

TABLE 2. Examples of included community-level and national- or policy-level interventions, classified by the components of the NOURISHING framework (19).

Intervention component	Examples of interventions	Comment
N 1	Regulations on standard nutritional information labelling; front-of-pack sugar warnings (24–26)	Least commonly reported
O	School nutrition policies; school lunch programmes; offering fortified snacks; healthier foods on supermarket shelves (23, 27–30)	Least commonly reported
U	Taxes and subsidy-framed messages on vending machines (26)	NA
R	Tackling marketing of breastmilk substitutes and general advertising to children (25, 31, 32)	NA
I 1	Reformulation of preschool meals; salt reduction by food businesses; wheat/rice fortification (21, 33, 34)	NA
S	Incentives for food businesses to reduce salt; school nutrition policies; taxation or import ban of unhealthy foods (21, 29, 35, 36)	Always accompanied by other components, most commonly with (O)
H	Legislative frameworks and intersectoral coordination to support food fortification; introduction of new cultivars to farmers; land reform; MAISMA project (WHO-guided multifaceted salt reduction intervention) (36–39)	Always accompanied by other components, most commonly with (I1)
I 2	Public awareness campaigns; promotion of national dietary guidelines; national policies to limit fat and salt intake; school nutrition programs (22, 29, 39, 40)	Commonly reported; typically ran in tandem with projects with wider scope, most commonly with (G) and (O)
N 2	Breastfeeding promotion; diabetes self-management programs (31, 41–44)	NA
G	School and community gardening; curriculum-based nutritional education within schools; curriculum-based behavior-change program targeting fruit and vegetable intake in middle-aged women (45–50)	Most commonly reported
Other	Provision of ready-to-use infant formulas in camps for internally displaced people during humanitarian response following an earthquake; training of children as peer influencers in school social networks to increase water consumption and decrease consumption of sugar-sweetened beverages (51, 52)	These interventions are not classifiable by the NOURISHING framework as they lacked sufficient detail or differed greatly from the existing components.

NA, not applicable.

Source: prepared by the authors from the results of this review.

Individual- and household-level interventions included: personalized breastfeeding promotion; provision of supplements, such as vitamins (especially vitamins A and D), minerals, or ready-to-use supplementary foods; and nutrition counselling.

Policy interventions

The WHO Noncommunicable diseases progress monitor 2020 (20) provides a definitive picture of nutritional policies in SIDS, of which 40 SIDS reported some policy information (Figure 2).

FIGURE 2. Proportion of Small Island Developing States (SIDS) fully, partially and not achieving nutritional policies.

Note: Eighteen of the 20 non-UN members/associate members of the United Nations Small Island Developing States listing were not included in the report, and contributed greatly to the “no data” component.

Source: prepared by the authors based on published data from World Health Organization Noncommunicable diseases progress monitor 2020 (20).

Across all SIDS regions, restrictions on the physical availability of alcohol (30 of 31 reporting SIDS) and increased excise taxes on alcohol (26 of 31 reporting SIDS) were the most commonly reported policies that were partially or fully achieved, while policies on saturated fatty acids and trans-fats were the least commonly reported (five of 40 reporting SIDS). Of fully achieved policies, restrictions on the physical availability of alcohol (seven of 31 reporting SIDS) and taxes on alcohol (eight of 31 reporting SIDS) were also the most commonly reported, while salt policies were the least commonly reported (two of 40 reporting SIDS).

Of the 40 reporting SIDS, Bahrain, followed closely by Maldives and Singapore (all SIDS in the AIMS region), took the lead on the number of partially or fully achieved policies (six of the seven policies). Guinea-Bissau reported the least number of policies achieved of SIDS in the AIMS region. Among Caribbean SIDS, Jamaica reported the most partially or fully achieved policies (four of the seven policies), namely physical availability of alcohol, alcohol taxes, and salt and fat policies. Antigua and Barbuda, Suriname, and Haiti reported the least number of policies achieved. Among Pacific SIDS, Timor-Leste reported the most partially or fully achieved policies (four of the seven policies), again on physical availability of alcohol, alcohol taxes, and salt and fat policies. Nauru, Niue, and Tuvalu reported the least number of achieved policies (Table 1A, supplementary material).

Examples of nutrition policies in the WHO Noncommunicable diseases progress monitor 2020 were described in some of the included records from the database search. With respect to fat policy, Fiji increased import duty on palm oil from 15% to 32% in 2012 (53). Samoa’s recognition of salt reduction as a priority action in its national food and nutrition policy and strategy for the prevention of noncommunicable diseases triggered its 2013 nationwide multifaceted salt reduction project, MASIMA (39). A range of salt-reduction activities were reported across Pacific islands, including working with the food industry to reduce sodium content of their food products and meals (25). Strategies for promoting breastfeeding in the Latin Caribbean, including the Baby-Friendly Hospital Initiative, length of maternity leave, and The International Code of Marketing of Breast-milk Substitutes were also discussed (31). Although an overall national strategy to restrict marketing of unhealthy foods and beverages to children is not listed for Fiji in the WHO Noncommunicable diseases progress monitor 2020, some packaged food, soft drinks and fast food companies are voluntary signatories to national or global industry initiatives (32). Policies on alcohol were not reported in any of the included records.

Barriers to and facilitators of interventions

Barriers to implementation and/or effectiveness of interventions ranged from typical constraints of project programming to more context-specific barriers within SIDS communities. High intervention costs were often cited and interventions on food provision were especially described as logistically complex and expensive. Cultural norms were a common community-level barrier; intervention contamination was often cited because of the culture of sharing, and sociocultural norms around food and preference for larger body size affected family acceptance of diet changes. Barriers related to the climatic, political, and economic vulnerabilities of SIDS also emerged. Climate factors in SIDS (e.g. natural disasters, heat, and humidity), socioeconomic downturns, and political upheaval reduced the availability of skilled human resources, study equipment, and healthy food options.

Facilitators of interventions were strongly reflected in relationships with stakeholders. Where researchers held an inclusive approach to working with stakeholders to ensure cultural relevance of interventions, strong relationships and a sense of ownership were formed, which translated into stakeholder support of activities and intervention success. A consequence of successful stakeholder collaboration was participant satisfaction with interventions. Researchers also attributed success to the use of multiple reinforcing components, such as the Marshall Islands Healthy Stores Program which incorporated five NOURISHING components (23).

More details of barriers to and facilitators of interventions are available in Annex 3 of the supplementary material.

DISCUSSION

This rapid scoping review found 174 unique interventions conducted in SIDS between 2000 and 2019. Most interventions were in the Caribbean region and were conducted at a national or policy level. Interventions tended to have multiple NOURISHING components, and the distribution of components was relatively consistent across SIDS regions. The most commonly reported components were giving nutrition education/skills (G), followed by informing about food/nutrition through public awareness (I²). Setting nutrition label standards/regulations on the use of claims and implied claims about food (N) and restricting food advertising and other forms of commercial promotion (R) were the least commonly reported.

NOURISHING

Framework and policy implications

It is plausible that the conventional provision of nutrition education/skills and promotion through public awareness were the most common intervention components because of the evolution of health promotion. Early days of health promotion were grounded in an individualistic model where people were thought to have more control of their health, and promotion activities sought to enable people to increase this control (54, 55). However, this model is shifting to include wider societal factors as significant influencers in the health choices individuals make (54, 55). For instance, the 2005 Bangkok Charter highlights the importance of determinants of health in a globalized world, encouraging action from a wider range of civil society, government, and international organizations to build policies and partnerships for health promotion (54). Our review found that these wide-scale, top-down types of interventions (for example, in the form of setting standards on nutrition labeling or restricting food advertising) were less commonly reported. Indeed, not only does setting legislation first require constant government commitment to population health, but managing the needs of many stakeholders with conflicting interests can be fraught with political implications (56). The private sector, particularly in the food industry, is a powerful stakeholder with considerable influence on national decision-making. Food and beverage companies profit from the sale of processed, unhealthy foods, and attempts at restricting these can be met with strong resistance. In 2021, the resistance of commercial interests affected the passing of a recent front-of-package warning label as a regional standard in CARICOM, whereby an unprecedented “emergency” revote of stakeholders in Jamaica contributed to the standard being rejected at the CARICOM level (57–59). Even so, long-term positive effects of dietary improvements in populations may not coincide with political and budget cycles (56). Thus, it is understandable why policy-makers might avoid these types of nutritional interventions and researchers may find difficulty advocating for them.

The WHO Noncommunicable diseases progress monitor 2020 illustrates the paucity of nutritional policies in SIDS, and their contextual factors offer a deeper understanding of this finding. The scarcity of advertising bans/restrictions on alcohol within the Caribbean and Pacific – in spite of many alcohol tax policies – can be partially understood by considering the historical and economic relevance of alcohol. Taxation is a blanket intervention applicable to many foods and services, with a key purpose of increasing national revenue (60). Given the economic vulnerabilities of SIDS, the simplicity and economic benefit of this type of policy make it desirable. Furthermore, physical restrictions on alcohol availability (for example, monopolization or licensing of retail outlets) can also increase national revenue, as outlets may be government run and/or licensing may require fees payable to the government. Alternatively, advertising bans and more comprehensive restrictions on alcohol likely hinge on deeper value systems. Large-scale production and exportation of alcohol during colonial times and the associated industrial prosperity have embedded alcohol in the society and culture of SIDS (61). Comprehensive restrictions on alcohol would not only affect these values, but also decades-old economic treaties of international trade in these regions (61).

Policies on saturated and trans-fat were the least commonly reported interventions achieved. As seen from the barriers reported in implementing interventions, populations may not easily support or comply with such policies as they conflict with cultural norms (56). Obesity was often cited by Pacific SIDS as a non-priority; in fact, it was seen rather as reflecting love, belonging and care, and a personal freedom separate from disease (27, 45, 62, 63). Similar norms are found in Caribbean populations where traditions of food preparation and consumption favor high-calorie ingredients and perceptions of body image favor larger body sizes (64–66). In addition, fat policy interventions that were enacted, such as bans on turkey tail imports in Samoa and mutton flap sales in Fiji (67), were wrought with political conflict between the ministries of commerce and health, individual stakeholders (mutton traders), and adherence to trade agreements (68).

Although considered an important means to improve nutrition (69), the concept of promoting the production and consumption of local foods was not commonly reported in the included records as an explicit way to improve nutrition. Barriers to this strategy are long-standing and complex. These include: international trade treaties in SIDS (including importation of foods); insufficient incentives for the local food chain with limited investment in commercial agriculture and improved agricultural technology; and growing urbanization and subsequent reduction in subsistence farming which relies on insufficient multisectoral rural development strategies (3). These factors may indicate why community food production initiatives were rarely examined in the context of diet-related outcomes (69).

Despite implementation challenges, the potential benefits of nutritional policies cannot be overlooked. The global success of policies on sugar-sweetened beverages highlights such potential; a 2019 Cochrane review concluded that environmental change – through labeling, restricting availability, taxation or other methods – led to reduced consumption of these beverages (70). This has been seen in SIDS such as Barbados where implementation of a 10% value-based tax was associated with decreased sales of sugar-sweetened beverages (71). Furthermore, combining multiple components of the NOURISHING framework to create a multipronged intervention may improve success by increasing its impact (72). National interventions on sugar-sweetened beverages implemented by multiple stakeholders and sectors are most effective, underscoring the importance of engagement and collaboration at all levels (70).

Reporting and implementation of interventions

Sufficiently detailed reporting of interventions is imperative for thoughtful and precise interpretation of findings and translation to other populations. This is especially relevant to SIDS whose interventions, often models from larger countries, must be tailored to suit their small-island context.

In tailoring interventions, it is crucial to recognize the factors that can affect their implementation and success. Working with stakeholders within an environment with climatic, political, and economic constraints can be considered context-specific to SIDS. Smallholder farmers are key to addressing nutrition insecurity in small islands, as is bridging formal and informal parties; yet, multistakeholder and multisectoral collaboration needs to be coupled with sufficient participatory action to be able to create an integrative food security policy that is both effective and acceptable (73). Merely stating the “use” of participatory action may not equate to the level of use that is required in the field. Adequate attention to each step of the collaborative process is needed; enhancing communication has been recommended by reviewing stakeholder relationships, identifying interdependencies, and reordering relationships, while allowing stakeholders to lead these processes (74). Systems dynamics methodology, which was not cited in any reports included, can assist in this refinement process and has shown success with wide-ranging types of participants through its use of participatory action for in-depth exploration of complex issues and/or systems (75). In any case, thoughtful adaptation of evidence-based models to specific populations, cultures, and contexts should be standard. The Diabetes Care in American Samoa project is a successful example of this (41); recognizing limitations in the amount of medical equipment and number of health professionals available in American Samoa, and the higher population blood glucose levels, researchers reframed their clinical algorithms to triage frequency and intensity of care. They also used an apprenticeship model with newly hired health workers to manage training costs and allow for culturally preferred collective working. When issues arose (e.g. sharing of equipment), there was effective communication between all stakeholders to come up with a culturally appropriate response. Where national policies are concerned, evaluations of government commitment should also be undertaken. This is effectively seen in G8 Research Group’s annual CARICOM monitoring grid for their commitment to the Port-of-Spain Declaration for the prevention and control of noncommunicable diseases and their risk factors in the Caribbean (76). This mechanism holds high-level stakeholders accountable to their commitments and encourages progress in each country, while allowing lessons to be learnt in the implementation of such commitments (76).

Strengths and limitations

This study was the first of its kind for SIDS and may inform regional policy for the implementation of food and nutrition interventions. However, as a rapid review, few databases were searched and neither was the grey literature. In addition, only reports in English were included (although few reports showed up in other languages), which likely reduced the number of relevant records captured in the literature search. In addition, publication bias favoring articles from Western or high-income countries means that research from many SIDS struggles to get published (77). Another limitation is that full details of an intervention may not be described in a given record, especially wide-scale policies. For example, an intervention that used economic tools to address food affordability (U) might have been accompanied by public awareness campaigns (I²), but the authors may not have discussed these campaigns in the manuscript. Thus, our results illustrate a partial snapshot of what evidence exists on nutritional interventions in SIDS and they must be interpreted within these confines.

Conclusion

There is a paucity of nutritional policies in SIDS; the reasons may be linked to their social, economic, and environmental vulnerabilities. Interventions should be expanded beyond simple education to encompass multiple components of the NOURISHING framework and promote multisectoral inclusion and thereby stronger stakeholder buy-in. A systematic review is warranted to examine a fuller range of sources and the effectiveness of interventions.

Disclaimer.

The authors hold sole responsibility for the views expressed in the manuscript, which may not necessarily reflect the opinion or policy of the Revista Panamericana de Salud Pública / Pan American Journal of Public Health and/or those of the Pan American Health Organization.

Funding Statement

The funder of this work, the International Development Research Centre, did not influence in any way the design of this study, or data collection, analysis, writing, and decision to publish these results.