Researchers be like “I know a caregiver”; it’s always the mother

Disclaimers: Views my own and I welcome feedback.

Just published a paper titled “Men’s nutrition knowledge is important for women’s and children’s nutrition in Ethiopia” where we looked at midline evaluation data from the ATONU trial, where chickens were given to small-scale poultry farmers in Ethiopia to improve women’s diet. We looked at men’s and women’s knowledge as measured by the FAO nutrition knowledge tool and looked to see if there were any associations with women’s and children’s dietary diversity. Key takeaway: there are.

I promise to make this post not just about my paper but how it fits into the larger picture of nutrition knowledge measurement and how nutrition in conceptualized. I had to cut 1300 words out of discussion, so this is definitely going to be in the context of other studies (I don’t know why we don’t have a supplemental discussion section – I am tempted to copy/paste these extra thoughts into “supplemental discussion” next time!)

Before I go on, this is a cross-sectional study (data from midline), and we do think this population (chicken farming households) may be better off than traditional farming households in Ethiopia, and we do agree there might be lot of assumptions on measurements that need further work (the focus of this post). We only collected knowledge starting at midline evaluation and collected knowledge on both men and women. We did briefly consider using a knowledge tool developed by Dr.Selamawit Bilal in Ethiopia but abandoned the idea because we wanted to collect something from both men and women that targets women’s diets (as well as child’s diets). Her work on identifying types of Ethiopian fathers: traditional, transition, and modern fathers, and how this affects their responsibilities and engagement with regard to child-rearing is really neat. (My quant brain thought of using latent class analysis to identify profiles of men/women who respond to nutrition education in LMIC – has this been done?).

And I don’t know everything so please kindly send me a note or comment if I have flawed interpretation or missed a sentinel publication.

Right now you may be asking:

1. What is nutrition knowledge and why should I care?
2. How do you measure it and what are you measuring?

I will address those questions here:

1.) Nutrition knowledge “refers to knowledge of concepts and processes related to nutrition and health including knowledge of diet and health, diet and disease, foods representing major sources of nutrients, and dietary guidelines and recommendation” (Miller and Cassidy, 2015). You should probably care about this because most chronic disease have dietary/nutrition causes and associations, not just for older people but for all life stages, especially as we see higher childhood obesity. So the idea is that if we increase nutrition knowledge of a target group, then they will make better food choice, and over time, this leads to better health outcomes.

From the LMIC (“Low and Middle Income Countries”) perspective, the advantage of improving “nutrition knowledge” is that it can “imitate” the effects of actual education received from schooling on nutrition outcomes. So among low-literate and low-resource population, nutrition education can help folks in making “nutritious” food choices (this way you can do more behavior change interventions without the larger structural interventions like schooling, which is substantially more costly, and time- and labor-intensive).

The majority of the nutrition knowledge education interventions in LMIC focus on women, who may not be ones who go to market or make decisions on food purchase (or have the money). I have seen a few studies that focus on grandmothers and some even on fathers – which is great, but let’s not forget siblings who might actually do the feeding of the younger child….so can we aim towards family-based nutrition education intervention in LMIC? (does this exist, is this effective? Please let me know!)

2.) Nutrition knowledge measurement: There is one tool that FAO uses to capture nutrition knowledge and this is based on Essential Actions in Nutrition, which is a really nice guide that gives an overview of nutrition goals and indicators for the world. I learned about the history of this tool from Schnieder and Masters, 2018 (open source, please read, highly recommend!). And just FYI: last year, Essentials Actions in Nutrition came out with a new publication that had a life-stages perspective, suggesting we should care about nutrition at all stages for all people (including men).

• Some history: FAO knowledge tool is based on a type of tool called KAP, which stands for Knowledge, Attitude, Practices, and the aim of this type of tool is to capture the situations and understanding of impediments to behaviors (apparently KAPs are from the 1950s and originated from the family planning world). In general, my sense is that they were popular 10-20 years back but have gone out of fashion as we explore more on what actually leads to behavior change. Now, people do barrier analysis or Trial of Improved Practices (TIPS), which to me seems like some version of KAP.[side-note] I don’t know why we don’t do a version of a tool that looks at knowledge enablers (market access, money, men, roads, food blender) rather than knowledge-barriers surveys (perhaps people do this through “positive deviance?”). More recent behavior change frameworks such as the COM-B framework includes “opportunity”, “motivation”, “capabilities” (Michie et al, 2011).
• Tool components: There are 13 modules available in the tool, ranging from diets of lactating and pregnant mothers, breastfeeding and early feeding of young children to knowledge on food safety, WASH, food-based dietary guidelines and even obesity! We used the 5 modules on diets on mothers, breastfeeding, early feeding, vitamin A and iron rich foods in our paper. We did expect high knowledge on breastfeeding compared to early feeding knowledge, because of couple reasons: 1) existing programs like Alive and Thrive and others in the regions, 2) but more importantly,  breastfeeding is one message and one source (mother), as opposed to buying, processing, cooking, feeding, diverse food groups such as eggs, meat, lentils, etc.
• Is this tool valid across all contexts? The FAO nutrition knowledge tool is field-tested in Cambodia, Malawi, Mexico and El Salvador. I am not sure if this tool is valid even across different populations within a country (urbanity/ages/education). In our paper, we mention “It’s possible the FAO instrument measuring knowledge is not adequate for capturing knowledge for men, or generally, for this context. In this case, we make the assumption that the measurement error with this instrument is similar between the gender, regions, and education.” We make this huge (and perhaps not fully justified) assumption because we don’t have an existing validated tool in Ethiopia. For farmers, for men and women, for low-literacy populations, for rural populations, for particular age groups, for these particular four regions (Amhara, SNNPR, Tigray, Oromia).
• New studies on nutrition knowledge: “Recent novel study in Malawi expanded nutrition knowledge to include food composition, future health and safety using foods that are available locally [Ex.which food contributes to future health? Nsima (maize meal) or Ndiwo (greens)?] (Schneider & Masters, 2018). They also found that behavioral knowledge has far greater effect compared to mechanistic knowledge similar to the results from this study. More importantly, the study highlights the need for next generation of contextualized nutrition knowledge instruments that are directly related to behaviors, country level dietary recommendations, and local food environment.” (more stuff I had to cut from discussion).
  
  Other studies have used Likert-type knowledge scales on what children should eat using local foods like snails in Nigeria and more generally, eggs and porridge in Ethiopia. There may be more, please let me know!

3.) Other thoughts / rants:

• Nutrition education on both men and women: In the discussion I talk about how Ragasa and colleagues (2019) “found that giving both men and women market access advice was significantly associated with higher household dietary diversity score (0.88 food groups), compared with men alone (0.31) or women alone (0.54) (Ragasa, Aberman, & Alvarez Mingote, 2019). In their study of 3001 households in Malawi, both members (women and men) received advice on market access in only 3% of households (Ragasa et al., 2019).” CAN YOU IMAGINE IF THEY WERE ABLE TO REACH MORE HOUSEHOLDS AND WHAT THE IMPACT MIGHT HAVE BEEN? Also, check out their really cool gendered framework of enabling pathways from agriculture to nutrition.
• Effect of nutrition knowledge on nutrition varies by child(ren), mother, father: Why don’t we look at women’s nutrition knowledge on her own outcomes in LMIC! (Note: women, not mothers; please read this excellent one-pager on how indicators on women are based on our reproductive organs).
• You know when you read a paper and there’s a solid statement that you agree with authors and you scream “THANK YOU”? Here’s one — “To the best of our knowledge, there are no studies assessing the role of women’s nutrition knowledge on her own dietary diversity or nutritional status in the developing world” (Schnieder and Masters, 2018).
  Also, can we start looking at men’s diets? And family diets?

• Small side-note on semantics: can we talk about how we refer only to the mothers as the “caregiver” but completely forget the father as a caregiver? I get there are female-headed households, but these are a small percentage. There are lots of studies that mention caregiver knowledge on child nutrition, but they always only refer to one parent. Researchers be like “I know a caregiver”, and it’s always the mother.

To conclude, we need research on the next generation of nutrition knowledge instruments for each country and target group (education, age, gender, rural-urban) in LMIC that are based on 1) foods promoted/recommended (not ‘healthy’) by the country’s dietary guidelines, and 2) preferably something that can also be tailored for family-based nutrition education. This ties into the larger issues of how nutrition or recommended diet is conceptualized, and how nutrition education should be operationalized. Maybe we can start by tailoring the FAO tool and borrowing lessons and marketing techniques from food industries within each country.

Final note: I came to all these “insights” after data collection, analysis, and literature review. Sometimes I have this fear that these papers are never good enough because I see so, so many weaknesses in measurements. Apparently, after talking to my peers and colleagues, this “paper paralysis” is normal. Anyways, I am trying to overcome this by being transparent about these findings, highlighting relevance (and lack of it) and documenting these digressions. Thanks for reading!