Behavior Change Communication Activities Improve Infant and Young Child Nutrition Knowledge and Practice of Neighboring Non-Participants in a Cluster-Randomized Trial in Rural Bangladesh

"Studies of nutrition BCC that do not account for information spillovers to non-participants may underestimate its benefits in terms of IYCN knowledge and practice."

This paper examines the impact on infant and young child nutrition (IYCN) knowledge and practice of mothers who were neighbours of mothers participating in a nutrition behaviour change communication (BCC) intervention in rural Bangladesh. The idea is that, where neighbours are a source of information on IYCN practices, BCC activities directed at mothers may generate information spillovers, with participants passing on this information to neighbouring non-participants. Drawing on qualitative fieldwork in rural Bangladesh, the researchers hypothesised that information provided as part of a high-quality nutrition BCC intervention was being passed on by participants to non-participants. They designed and implemented a quantitative study to address this hypothesis in rural Bangladesh, where stunting remains widespread and IYCN practices sub-optimal.

The researchers used data from two, 2-year, cluster randomised control trials (RCTs) that included nutrition BCC in some treatment arms. Conducted between March 2012 and May 2014, the Transfer Modality Research Initiative (TMRI) was designed and evaluated by the International Food Policy Research Institute (IFPRI) and implemented by the United Nations' World Food Programme (WFP). Two RCTs were conducted in rural areas: In the north, study villages were randomly assigned to a control group or one of four treatment arms in which beneficiaries received a cash transfer ("Cash"), a food ration ("Food"), a half cash transfer and half food ration ("Cash&Food"), or a cash transfer along with nutrition BCC ("Cash + BCC"). In the south, study villages were also randomly assigned to a control group or one of four treatment arms; the first three treatment groups were the same as in the north. The final treatment group in the south was different: Instead of a cash transfer along with nutrition BCC, beneficiaries received a food ration along with nutrition BCC ("Food + BCC"). All beneficiaries were economically poor households with a child aged 0-24 months in March 2012.

The core BCC activity consisted of a weekly, one-hour group session of the 10 beneficiaries in each village with a trained community nutrition worker (CNW). These sessions covered: nutrition, diet diversity, and health; handwashing, hygiene, and health; diet diversity and micronutrients; breastfeeding; complementary foods for children 6-24 months; feeding and treatment of children with diarrhoea; maternal nutrition; encouraging homestead food production; and women's status and relationships with influential family members (particularly husbands and mothers-in-law) and the wider community. A variety of methods were used to deliver this information, including presentations, question and answer, interactive call and answer, songs and chants, practical demonstrations, and role playing. Some sessions were held exclusively for beneficiaries; for others, husbands, mothers-in-law and other influential individuals from beneficiaries' homes were encouraged to attend. In addition, CNWs made home visits to beneficiaries to follow up on topics discussed during the group sessions and to discuss specific concerns that mothers might have. When a mother missed a session, the CNW followed up with a home visit. In addition, CNWs and staff fro the Eco-Social Development Organization (ESDO), a non-governmental organisation (NGO) contracted by WFP, conducted meetings with influential members (for example, village heads, religious leaders, school teachers, and local health and family planning staff) of the villages in which the BCC took place to explain the purposes of the nutrition training and to provide them with the information being conveyed to study participants. CNWs received training prior to the start of the intervention, with refresher training undertaken three and 12 months after the intervention began.

A survey of TMRI neighbours in both the north and south was conducted in April 2014 during the final month of transfer payments, concurrent with the TMRI endline survey. These neighbours were not participants in either randomised trial. A "BCC neighbour" household: (a) is a neighbour of a TMRI household that had received nutrition BCC in addition to their cash or food payment; and (b) had a child aged 0-24 months at the time of the survey. A "non-BCC neighbour" household: (a) is a neighbour of a TMRI household that had received a cash or food payment but no BCC; and (b) had a child aged 0-24 months at the time of the survey.

The researchers analysed data from 300 BCC neighbour mothers and 600 non-BCC neighbour mothers. They constructed measures capturing mothers' knowledge of IYCN and measures of food consumption by children 6-24 months. (This was based on seven questions assessing mothers' knowledge of optimal breastfeeding practices and an additional seven questions on knowledge of complementary foods, foods important for micronutrient intake, and hygiene.) The effect on outcomes of exposure to a neighbour receiving a nutrition BCC intervention was estimated using ordinary least squares and probit regressions.

The researchers also asked mothers in our neighbour sample about the sources of information on IYCF available to them. They listed a wide variety, including family members, health workers, the media, NGOs such as BRAC, and neighbours, including those neighbours who are recipients of nutrition BCC through TMRI. 35% of these mothers reported receiving information from neighbours. Only health centres were reported as frequently as a source of information (35%), with the next most frequently reported sources being family (21%) and BRAC (21%). They also asked the mothers in the neighbour sample if they were aware of critical IYCF practices. Awareness of these was high; for example, 93% of mothers reported that they had heard about the importance of starting breastfeeding immediately after delivery or within one hour.

The researchers found that having a neighbouring mother participate in the nutrition BCC intervention increased non-participant mothers' IYCN knowledge by 0.17 standard deviation, or SD (translating to 0.3 more correct answers). This effect is statistically significant (P = 0.04). Relative to non-BCC neighbors, BCC neighbours were: 14.1 percentage points more likely to feed their 6-24 months children legumes and nuts; 11.6 percentage points more likely to feed these children vitamin-A-rich fruits and vegetables; and 10.0 percentage points more likely to feed these children eggs. Children of non-participant mothers who had a neighbouring mother participate in a nutrition BCC intervention were 13.8 percentage points more likely to meet World Health Organization (WHO) guidelines for minimum diet diversity, 11.9 percentage points more likely to meet WHO guidelines for minimum acceptable diet, and 10.3 percentage points more likely to meet WHO guidelines for minimum meal frequency for children who continue to be breastfed after age 6 months. Children aged 0-6 months of non-participant mothers who are neighbours of mothers receiving BCC were 7.1 percentage points less likely to have ever consumed water-based liquids.

Nesting this work in the context of a randomised trial gives the researchers confidence that these effects are causal and not just associational. The study, however, has weaknesses. For example, the precise mechanism underlying the spillover effects is unclear. While the qualitative data tells us that mothers in participating households talk with other mothers, the TMRI intervention also included activities such as community meetings. It is possible that these meetings, and not informal mother-to-mother interactions, may have contributed to changed behaviour and practice among non-TMRI participants. However, the community meetings were primarily with influential community members such as village heads and religious leaders, rather than with non-TMRI mothers in the community. Moreover, 98% of BCC participant mothers in the endline survey reported passing on the information to other people.

One implication of these results is that existing studies of the impact of nutrition BCCs may underestimate their benefits: Existing studies focus only on participants in these interventions, yet the above-summarised results show that non-participants may also benefit.