Money or Knowledge? Behavioral Aspects of Malnutrition

Weiss Family Program Fund, Harvard Center for African Studies, Harvard Foundations of Human Behavior, Vogelheim Hansen Fund

Malnutrition accounts for nearly half of child deaths worldwide. Children who are well-nourished are better able to learn in school, grow into more physically capable adults, and require less health care during childhood and adulthood. Moreover, it is difficult to make up for poor childhood nutrition later in life. I present here the proposal for an intervention that builds on a larger study in Ethiopia and will generate insights into the importance of behavioral factors related to persistent malnutrition in low-income settings, allowing for more targeted, cost-effective interventions in the future. Existing data from the study region, Oromia, Ethiopia, suggest that many mothers know how to correctly respond to a hypothetical situation where a young child exhibits poor growth. On the other hand, however, mothers frequently appear unaware about their own children's growth deficiencies. Together, these facts suggest that false beliefs about the appropriateness of a child's physical size are a more likely contributor to malnutrition, rather than a weak understanding of how to help a malnourished child. The proposed intervention will provide evidence on the relationship between caregiver beliefs about child nutritional status and the caregiver's behavior, ultimately analyzing how this relationship influences important nutritional choices for young children in a setting with limited resources. The study uses a two-by-two randomized trial; the first treatment is a cash transfer labeled for child food consumption, and the second is the provision of personalized information about the quality of the child's height compared to other children like those of the same age and gender in East Africa. Together the two treatment arms will provide evidence about the relative importance of behavioral versus resource barriers to improved nutrition. Better understanding of the interaction between these key factors is essential in addressing one of the foremost health issues facing developing countries today.

All households in the study were given general child nutrition educational messaging, immediately after the baseline survey and prior to any treatments. This messaging focused on appropriate feeding habits complemented by breastfeeding and ways to maintain proper hygiene during food preparation and consumption.

The household received both the personalized information intervention and labeled cash transfer intervention.

During a prior study in June-July 2016, we collected anthropometric measures on the index children, including the children's height. Based on these data, for households assigned to the information treatment, enumerators provided personalized information to the children's primary caregiver about the index child's current height, during a baseline household visit. The enumerators carried a display card that visually showed where the child's height fell compared to "healthy" children of the same age and gender like those in East Africa. The enumerators emphasized to the caregivers that short stature is due to poor chronic malnutrition and is not just attributable to genetics or a recent illness. During this visit, the enumerators additionally pointed out that chronic malnutrition is not immediately life-threatening.

Households received a cash transfer labeled for child food consumption and were told the money is designed to cover additional spending for food for the index child (and any other younger children in the household) over the next six weeks. Though it was given as a single, lump sum payment, the transfer was evenly split and handed to the household in six sealed envelopes, to help the households better allocate the money. To further encourage them not to spend the money all at once, each envelope was labeled with a number, the index child's name, and the dates for the week the money in the envelope should be spent. Enumerators clearly stated that this is a one-time money transfer.

Number of foods that index child consumed in past 24 hours from among: grains, tubers, milk, vitamin-A rich fruits and vegetables (e.g., pumpkins, carrots, dark leafy vegetables, mangoes, papayas), other fruits and vegetables, animal protein foods, and legumes, as measured through an interview with the child's caregiver at 6 weeks post intervention

Number of days in past week that index child consumed key foods (meat/fish, fruits, vegetables, eggs, milk/dairy products, legumes), as measured through an interview with the child's caregiver at 6 weeks post intervention

Number of times child was fed in previous 24 hours; assessed separately depending on whether child is still breastfeeding, and by age group (<24 months, 24-36 months, >36 months), as measured through an interview with the child's caregiver at 6 weeks post intervention

Total score from: Dietary diversity (0 or 1 foods = 0 points, 2-3 foods = 1 point, 4+ foods=2 points), food frequency (0 days = 0 point, 1-3 days = 1 point, 4+ days = 2 points), breastfeeding (1 point; relevant for children up to 36 months), and meal frequency (0-1 meals = 0 points, 2 meals = 1 point, 3 meals = 2 points, 4+ meals = 3 points), as measured through an interview with the child's caregiver at 6 weeks post intervention

Household spending on key foods (meat/fish, fruits and vegetables, eggs, milk/dairy products, legumes)

Households for this study were selected from among those who were included in any of the three study groups from a larger study and for whom relevant data had been collected. Inclusion Criteria: inclusion in the larger study required the household to have a child who was 6-35 months old for the main study's baseline survey in July-August 2015 (referred to as the index child) and for the household to have land for crop cultivation Exclusion Criteria: Households that did not meet the inclusion restriction or those who did not have anthropometric data collected during the larger study.

Ruel MT, Menon P. Child feeding practices are associated with child nutritional status in Latin America: innovative uses of the demographic and health surveys. J Nutr. 2002 Jun;132(6):1180-7. doi: 10.1093/jn/132.6.1180.

Arimond M, Ruel MT. Dietary diversity is associated with child nutritional status: evidence from 11 demographic and health surveys. J Nutr. 2004 Oct;134(10):2579-85. doi: 10.1093/jn/134.10.2579.