Skip to main content

Accessing diverse diets: age and gender in household food allocation

Kathryn Merckel is a Research Support Specialist and former Summer Intern with the TCi program, and has just completed her Masters in International Development. She is interested in the intersection of nutrition and gender, particularly empowerment, behavior change, and child care concepts. She will be starting a PhD in Nutrition this fall at Cornell, where she will continue to work with the TCi team.

This year, I have been working on an exciting new way to use information about dietary diversity. In India, several studies have found that all members of a household may not have the same access to food. Some members may have preferential access to highly nutritious foods such as spinach and eggs, while other members go without, eating a calorie sufficient but low nutrient diet of starchy staple foods. If this is the case in a community where you work, it may mean that even though households have enough food to meet the needs of everyone in them, some people may be losing out and be malnourished. Often, these people are the most vulnerable already: women, children, and the elderly. In the figure below (from Jones, A., Ngure, F., Pelto, G., & Young, S. (2013)) we see how this “intra-household food allocation” fits in the pathway from overall availability of food to individual nutrition. Because of this linkage, knowing the degree of inequality among household members when it comes to food consumption is vital in designing programs that will reach those most in need.

image

From from Jones, A., Ngure, F., Pelto, G., & Young, S. (2013)

Unfortunately, it is very difficult to measure how food is distributed among household members. Humans are notoriously bad at remembering details correctly and have a very tough time estimating amounts of food they have consumed. Do you remember what you ate for lunch yesterday, let alone last week? Observing a family at meal times may influence their behavior, particularly in resource-poor areas or where there are strong cultural elements in the preparation of food. A family might feel compelled to cook a better meal than usual because you are a guest, or cook more food than they normally would so that they can share it with you. It is for these reasons that the dietary diversity survey was developed. It requires an individual to remember only the foods they ate the day before, and does not take into account the amount of foods consumed, only the type. Scores of studies have shown that a diverse diet is a predictor for better nutrition and health.

DDdiff: Dietary diversity differences within the household

Because dietary diversity surveys can obtain information on diet quality for every member of a household, it is an ideal place to look when trying to understand the intra-household food allocation patterns of a particular village. I used a dataset collected by the International Crops Research Institute for the Semi-Arid-Tropics (ICRISAT) in 2013-2014 that included dietary diversity scores for every member within the sampled households. I found the average dietary diversity score for each household, then for each individual in the household, found the difference between their dietary diversity score and their household average score. This then provided a measurement of an individual’s diet quality relative to the other household members. I called this score the “Dietary Diversity difference” or DDdiff. Individuals with a negative DDdiff consumed a diet less diverse than the average of their household, while individuals with a positive DDdiff consume a more diverse diet than their household on average.

It is thought that intra-household food allocation discrimination is very regionally specific: some villages may have many households where women eat a very limited diet, whereas in other places there is equality among adults but girl children have less access to nutrient rich “high value” foods than boy children. For this reason, we have to assume that patterns of intra-household food allocation apply only to the area in which they are studied. Having a relatively easy mode of detecting these patterns is important, as it can indicate presence of discrimination in a particular village, alerting the researcher to an issue that requires more attention. There is also evidence that the severity of any discrimination may fluctuate based on the availability of food and food prices. In this method, I controlled for both village and season.

Surprising results: The importance of age

I found some very interesting results. For these data, I found that the biggest effect in predicting an individual’s difference from their household mean was not gender, but rather age. Children and youth were significantly more likely to have a diet more diverse than their household average, while adults aged 36-55 were significantly more likely to have a diet less diverse than the household average. The only age group where gender was a mediating factor was in the 26 to 35 group, where women are predicted to have a diet significantly less diverse than the household average compared to their male counterparts. For women, this may be an age that they are marrying, moving out of their natal homes, and into a new household. They are also likely to be having children at this age. This makes them particularly vulnerable, both sociologically and physiologically.

image

Young women who have recently moved into their husbands home may be more sociologically and physiologically vulnerable
(Photo Credit: Jessica Ames, Ranchi, Jharkhand. February 2015)

The importance of these findings

While this method has its shortcomings, and only provides a rough estimate of the discrimination that may be occurring within households, it is a great step towards understanding what factors are at work in mediating malnutrition. With refinement, this method could prove to be a useful tool in designing programs aimed at reducing malnutrition, particularly in women and children. In the future, I hope to look further into factors such as education level, household structure, and marriage status to better understand which individuals are most at risk of being food insecure while living in a food secure household.