How do nutrition professionals working in low‐income countries perceive and prioritize actions to prevent wasting? A mixed‐methods study

Abstract Despite a shared commitment to achieving global nutrition targets, development and emergency‐humanitarian actors tend to prioritize different nutrition outcomes and actions. New approaches are needed to bridge the divide between these communities and to strengthen the overall evidence base for prevention of wasting. To better understand how these different groups perceive and prioritize actions for wasting prevention, key informant interviews (n = 21) were conducted, and an online survey was fielded among nutrition professionals working in low‐income countries (n = 107). Additionally, nutrition policy and strategy documents for select global and country institutions (n = 12) were analysed to identify interventions and approaches for addressing different forms of undernutrition. The findings of this study suggest that at both global and country levels, development actors tend to prioritize stunting prevention, and emergency‐humanitarian actors tend to prioritize treatment of acute malnutrition. It was less common for wasting prevention to be mentioned as an explicit priority. Many interventions were perceived by respondents to influence both stunting and wasting despite limited published evidence of effectiveness on wasting for water, sanitation and hygiene, growth monitoring and promotion, breastfeeding promotion and micronutrient supplementation. To help unify the nutrition community around prevention of wasting, the discourse about priority interventions should shift from ‘stunting versus wasting’ and ‘prevention versus treatment’ to a life‐course framing around child survival, growth and development. Respondents identified a need for more programme and research funding that prioritizes both wasting and stunting as outcomes. They also suggest leveraging existing national coordination bodies that bring development and emergency‐humanitarian partners together.

ties and to strengthen the overall evidence base for prevention of wasting. To better understand how these different groups perceive and prioritize actions for wasting prevention, key informant interviews (n = 21) were conducted, and an online survey was fielded among nutrition professionals working in low-income countries (n = 107).
Additionally, nutrition policy and strategy documents for select global and country institutions (n = 12) were analysed to identify interventions and approaches for addressing different forms of undernutrition. The findings of this study suggest that at both global and country levels, development actors tend to prioritize stunting prevention, and emergency-humanitarian actors tend to prioritize treatment of acute malnutrition. It was less common for wasting prevention to be mentioned as an explicit priority. Many interventions were perceived by respondents to influence both stunting and wasting despite limited published evidence of effectiveness on wasting for water, sanitation and hygiene, growth monitoring and promotion, breastfeeding promotion and micronutrient supplementation. To help unify the nutrition community around prevention of wasting, the discourse about priority interventions should shift from 'stunting versus wasting' and 'prevention versus treatment' to a life-course framing around child survival, growth and development. Respondents identified a need for more programme and research funding that prioritizes both wasting and stunting as outcomes. They also suggest leveraging existing national coordination bodies that bring development and emergency-humanitarian partners together.

K E Y W O R D S
acute malnutrition, interventions, nutritional status, stunting, wasting prevention

| INTRODUCTION
A range of interventions for the prevention and treatment of malnutrition in young children is being scaled up across countries with high burdens of undernutrition (Bhutta et al., 2013). Despite generalized support for combatting malnutrition, there has been 'fragmentation of interests and perspectives' within the global nutrition community related to which nutrition outcomes different camps of actors prioritize (Menon & Stoltzfus, 2012). Historically, childhood wasting-often called acute malnutrition-was the global picture of malnutrition.
Images of famine in Ethiopia in the 1970s and 1980s (BBC, 2014) reinforced wasting as the predominant view of malnutrition, and it continues to be the main priority of the emergency-humanitarian community. However, stunting-also referred to as chronic malnutrition-overtook wasting as the dominant priority of many development-focused actors with the release and related advocacy around the 2008 Lancet Series on Maternal and Child Undernutrition (Bhutta et al., 2008;Leroy & Frongillo, 2019;Victora et al., 2008).
These two groups, emergency-humanitarian and development, differ in the frameworks used to characterize and respond to malnutrition problems (Wasting and Stunting Technical Interest Group, 2018). Stunting reduction efforts concentrate on prevention interventions among children under 2 years while wasting-focused strategies prioritize screening and treatment of wasted children under 5 years using community-based protocols (Kennedy, Branca, Webb, Bhutta, & Brown, 2015;Stewart, Iannotti, Dewey, Michaelsen, & Onyango, 2013).
The prevention of wasting in particular has received less attention than the prevention of stunting; however, the situation is evolving. (WaST) Technical Interest Group, which has advocated that more research and interventions need to be focused on identifying and addressing the common environmental and dietary risk factors and physiologic processes that contribute to both ponderal and linear growth faltering (Angood, 2014). These two conditions often co-exist in the same populations of children (Khara, Mwangome, Ngari, & Dolan, 2018;Schoenbuchner et al., 2019;Uauy, Garmendia, & Corvalán, 2014;Wells et al., 2019).
Given common risk factors, it seems logical to assume that many interventions to prevent stunting will also prevent wasting (Briend, 2019;Myatt et al., 2018;Walson & Berkley, 2018). However, there is a complex and often oversimplified etiological relationship between wasting and stunting that still needs to be elucidated (Briend, 2019;Schoenbuchner et al., 2019). A recent review by ENN found that the current evidence base is not sufficient to conclude which interventions prevent wasting in non-emergency contexts (Emergency Nutrition Network [ENN], 2019). In a review of evidence carried out for The Lives Saved Tool, 2020 (Panjwani & Heidkamp, 2017), members of our research team found that fewer studies of nutrition education interventions reported on wastingrelated outcomes as compared with food supplementation interventions. They also found that ponderal outcomes were less consistently reported across studies compared with linear growth. (Panjwani & Heidkamp, 2017). This raised questions about whether the variability in reported outcomes across studies also reflected differing priorities of development versus emergency-humanitarian actors behind the studies (Ickes, Craig, & Heidkamp, 2019). To achieve global targets, new approaches are needed to bridge divides and to strengthen the overall evidence base for wasting prevention. Most policy and programme decision-making is based on the perceptions of stakeholders (Pelletier, Menon, Ngo, Frongillo, & Frongillo, 2011;Tumilowicz et al., 2019

Key messages
• Ongoing divisions of the nutrition community between development and emergency-humanitarian actors hinder the prioritization of wasting prevention at global and country levels.
• Despite limited or weak evidence for their effectiveness on wasting, many nutrition professionals believe that water, sanitation and hygiene; growth monitoring and promotion; breastfeeding promotion; and micronutrient supplementation will influence wasting.
• For promoting wasting prevention, a life-course framing that emphasizes child survival, growth and development is likely more effective than an outcome-focused framing on stunting and/or wasting.
• Programme implementation and research initiatives as well as their funding institutions should include targets and metrics for both wasting and stunting. online stakeholder survey and an analysis of nutrition-focused strategic documents from organizations that work in low-and middleincome countries (LMICs).  (Panjwani & Heidkamp, 2017) and a recent review of literature that examines the relationships between wasting and stunting (ENN, 2019). Focal countries were purposively selected from two high-burden regions (sub-Saharan Africa and Asia) based on a relatively high prevalence of stunting and wasting (Table SS1) and funder priorities. The sampling method was purposive in order to obtain information-rich cases with expertise around nutrition policies and programmes in low-resource contexts (Patton, 2002).
The selected focal countries reflect the high rates of stunting and wasting in sub-Saharan Africa and Asia, the two regions with the highest global burden (Alderman, Behrman, Glewwe, Fernald, & Walker, 2017). The authors adhered to the Consolidated Criteria for Reporting Qualitative Research (Table S3) (Tong, Sainsbury, & Craig, 2007). Eligible participants were identified by the researchers through professional networks, which included both academic researchers and key personnel at nongovernmental organizations (NGOs) working within or across the focal countries as well as snowball sampling within recruited respondents. Thirty eligible participants were recruited via email. Twenty-one (70%) participated, and nine did not respond; however, none directly refused to participate. In one instance, two individuals from the same organization were interviewed.
The lead author (S. I.), a professor of nutrition with expertise in maternal and child nutrition in low-income contexts, conducted all KII via phone in English. The interviewer had a prior collaborative relationship with two of the 21 KII participants.
The average interview duration was 40 min. Interview participants were sent the interview guide and details on researcher's goals for the project via email in advance of the interview. Verbal informed consent was obtained prior to starting the interview. The interviews were audio recorded and transcribed verbatim. No repeat interviews were conducted. Interview transcripts were analysed in Dedoose qualitative analysis (Los Angeles, CA, 2018) by four trained research assistants using thematic codes. The research team developed the codebook collaboratively to ensure consistent code definitions were understood by all coders. The first two interviews were used as a test case to ensure that an interrater reliability of at least 90% was achieved. In the rare case of a coding ambiguity, researchers consulted with the lead author (S. I.). All interviews were double coded to ensure consistency of data classification.
To illustrate key themes, rich text quotes were extracted along with de-identified participant information.
A combined grounded theory and content analysis approach was used; codes were developed deductively from the interview guides and inductively from interview findings (Miles, Huberman, & Saldaña, 2014). Code reports were produced for each code to be analysed, and were then grouped according to five main topic families associated with the research questions. Fifteen initial codes (- Table S4) were developed that were further grouped into four broader categories: (a) intervention types and categorization; (b) treatment of MAM and prevention of wasting; (c) factors used to target nutrition interventions; and (d) recognition of nutrition intervention 'camps' and ideas for building harmonization.

| Stakeholder survey
Following the KII, an online survey was developed and disseminated using Qualtrics (Provo, UT, 2018) from April to May 2018. The aim of the survey was to capture the perspective of a wider range of respondents working across LMIC. Participants were invited to complete the survey using a mass email distribution or personal email invitation from the authors. Emails were sent to the following listservs: Accelerated Reduction Effort on Anemia, Ag2Nut, Core Group Nutrition and USAID SPRING Nutrition Budget and Planning. In addition, the 21 individuals who participated in KII were invited to participate in the online survey. Online survey recruitment began in late April and was closed in mid-May 2018.
The online survey was developed from the findings of the KII and asked specifically which nutrition interventions are implemented in the respondent's context and whether the divisions described in the report introduction were apparent in their context.
In addition, using a list of 20 interventions identified through KII, respondents identified which outcomes (wasting, stunting, both wasting and stunting or neither wasting nor stunting) they believed each intervention would influence. Respondents were asked an open-response question about meaningful ways to categorize interventions. For the question about whether stunting and wasting are addressed by the same project or organizations in their context, participants were given three response options (agree a little, agree somewhat and agree a lot) but could not select 'do not agree' , which was an oversight in survey design.
The survey was conducted in English. No sample size target was determined a priori, rather the researchers sought to obtain the largest sample possible within the specified time period. Participants provided written electronic consent; however, responses could not be linked to individual respondents due to lack of identifiers.

| Ethical considerations
Wheaton College and Johns Hopkins University School of Public Health Institutional Review Boards approved the study procedures.

| RESULTS
Twenty-one KII were conducted between January and March 2018.
Five participants were employed in UN agencies, nine were from NGOs, four were from governmental organizations and three were nutrition researchers employed at universities. Eleven of the participants were nationals of one of the focal countries. Five KII participants were primarily engaged in emergency-humanitarian organizations or research, while the remainder worked for organizations that spanned both development activities and emergencyhumanitarian activities.
A total of 107 respondents participated in the online survey, of whom 75 completed all questions. The majority of respondents worked for NGOs (53%, n = 56), followed by government (16%, n = 17) and university or research institutions (18%, n = 20). Common professional responsibilities included programme planning, daily programme management, research and programme evaluation. About two-thirds (64%, n = 68) of online survey respondents worked in a single country, and one-third (36%, n = 39) worked across multiple countries. The three most common regions for work were sub-Saharan Africa (44%, n = 47), South Asia (34%, n = 37) and Latin America/Caribbean/South America (16%, n = 17).

| Perceptions and prioritization of wasting
Findings from all three data collection activities-KII, online survey and review of strategic documents-suggest that prevention of stunting is generally a higher priority objective than wasting prevention.
Many KII national respondents said that wasting remains a critical nutrition problem. However, several suggested that global efforts to prioritize stunting prevention may contribute to de-prioritization of wasting in their contexts. As one participant noted, 'Some donors see stunting as a "bigger picture" than wasting' (Tanzania). Another respondent noted: '[The prevention of stunting] has become a big, big objective in the world' (Burkina Faso). Some participants noted that the lower prevalence of wasting may also reduce the incentive to address it: 'Your reach in a wasting program is going to be a lot lower.
Because not as many people are affected, it tends to be deprioritized a little bit' (Multi-country stakeholder, United States).
While participants understood the term 'wasting prevention' and felt it would be understood in their contexts, they said it was not commonly used.
One intervention I can think of on the health side is community case management. This kind of a preventive approach because you are trying to quickly get to a child at the community level with treatment for an illness which would then prevent the wasting that would come probably right after that incidence of illness. So, I  Our findings did not indicate any notable differences in views regarding wasting and stunting by geographic region or working within a single versus multi-country setting.
The strategic document review highlighted differences in prioritization of approaches for wasting and stunting across government and development partner strategies. This analysis, summarized in Table 1, suggests that wasting prevention is a named priority, but it lags behind both wasting treatment and stunting prevention. Six institutions explicitly identified the prevention of wasting as a goal compared with eight that identified treatment of moderate acute malnutrition (MAM) and eight that identified prevention of stunting.
Notably, among the 10 strategies that included monitoring and evaluations plans, stunting and wasting were both included as key performance indicators in seven plans. KII also showed that stakeholders associate certain interventions like IYCF promotion with influencing multiple outcomes:

| Views on effective interventions for wasting
We are stunting focused, but address wasting and the most common micronutrient deficiency: anaemia.
When you make a very good promotion of IYCF practices, I think you have impact on stunting reduction, anaemia reduction, and also wasting reduction (Singlecountry stakeholder, Burkina Faso).
In For the micronutrient interventions there were notable differences between respondents who worked across multiple countries (n = 27) compared with those working in a single country (n = 48).
Single-country respondents were more likely to identify multiple micronutrient supplements (MMS) during pregnancy as having impact on both stunting and wasting (45%, n = 22 of single-country versus 19%, n = 5 of multi-country). More than half of multi-country respondents (54%, n = 15) selected 'neither outcome' for MMS compared with 13% of single-country respondents (n = 6). There was a similar pattern of selecting 'neither outcome' for iron and folic acid supplementation during pregnancy (62%, n = 17 of multi-country vs. 23%, n = 11 of single country) and for high-dose vitamin A supplementation in children (64%, n = 17 of multi-country vs. 23%, n = 11 of single country). The subgroup differences may be attributable to multicountry stakeholders from UN agencies, donors or multinational NGOs having more access to global evidence reviews than those focused on single countries who are more likely to work for government or local NGOs.

| Perceptions of SAM versus MAM treatment
Some KII saw the potential for 'treatment of MAM' to prevent SAMbut several considered this an impractical and unsustainable strategy given the low coverage and high cost relative to the burden. The flaw in that kind of thinking is that, the challenge is that stunting is a multifaceted process and the interventions that are going to reduce it will have to be multifaceted. So, that is kind of my view of the landscape (Multi-country stakeholder, United States).
The online survey findings indicated a lack of strong agreement about whether the same projects and organizations address both stunting and wasting in stakeholder contexts. Only 35% agreed strongly with the statement that 'the same projects address both stunting and wasting' (n = 37, 35%). Fewer (n = 22, 21%) agreed strongly that 'the same organizations are responsible for planning and implementing projects to address both wasting and stunting.'

| Strategies for bridging divisions within the nutrition community
Stakeholders suggested different approaches for bridging the divide between development and emergency-humanitarian groups. Those working across multiple countries advocated for greater harmonization in funding sources and reframing discourse and terminology around wasting. Respondents working in single countries emphasized improvements in national coordination.

| Reframing discourse and terminology
Within the organizational strategic plans, the most common classifications used to describe intervention activities were using the life course (pregnancy, infancy, early childhood and adolescence) and 'nutrition sensitive versus nutrition specific' (Table 1).
Among survey participants, 41 suggested an alternative categorization to 'stunting versus wasting focused' in response to an open-ended question. The most common suggestions were life course or age-based (n = 11, 27%), 'prevention versus treatment' (n = 9, 22%) and by the F I G U R E 1 Online survey respondents' categorization of maternal and child nutrition interventions by outcome (n = 107). BMI, body mass index; SAM, severe acute malnutrition; WASH, water, sanitation and hygiene nutritional status being targeted (e.g., micronutrient deficiency and chronic undernutrition) (n = 9, 22%). Participants also suggested grouping interventions according to whether they achieve intermediate or long-term outcomes (n = 2, 5%), the delivery platform of the intervention (n = 1, 2%), or more generally 'targeting' (n = 8, 20%).
Integration of services and harmonization of priority outcomes towards the common goal of a well-nourished or healthy child was a repeated theme. Reframing the discourse was seen as a strategy to The need for stunting and wasting to both be prioritized by governments, donors and implementing partners is echoed by a recent commentary by Leroy and Frongillo (2019) that critiques the prioritization of stunting. The authors advocate for using outcomes that are more proximal to interventions and inclusive of other dimensions of child wellbeing.

| Strengths and limitations
The primary strength of this study is the application of multiple methods to obtain the perspectives of a wide range of nutrition stakeholders including those working within single countries and across multiple countries, who often have different priorities. This study also had several limitations. First, the online survey did not require responses to all questions and therefore had limited responses to select questions. Second, the study may have been affected by selection bias due to non-response, whereby more motivated or invested respondents chose to participate. Individuals with lower access to the internet may have been less likely to participate in the online survey or a Skype interview. In an effort to mitigate telecommunications issues, interviewees were called by phone directly when platforms such as Skype were problematic for participants, and interviews were scheduled during the most convenient times for the participants. Both KII and online surveys were only conducted in English, which may have biased the responses received. Third, information bias may have been introduced due to the omission of a 'do not agree' option for the two survey questions that asked if the same projects and organizations address both stunting and wasting in participant contexts. While bias is possible from this design oversight, if respondents truly did not agree, the logical alternate selection would have been to 'agree a little' versus to 'agree somewhat' or 'a lot' . Given that the majority of participants did not 'agree a lot' with these two questions, any information bias that resulted from this design flaw appears to be minimal.

| CONCLUSION
The burden of multiple forms of malnutrition remains high in many countries. One in five children is stunted (Global Nutrition

CONFLICTS OF INTEREST
The authors declare that they have no conflicts of interest.

CONTRIBUTIONS
SI, CC and RH performed the research. RH and SI designed the research study. SI, RH and CC analysed the data and wrote the paper. All authors read and approved the final manuscript.