An emergency cash transfer program promotes weight gain and reduces acute malnutrition risk among children 6-24 months old during a food crisis in Niger

Abstract

Background

Assessment of the impact of emergency cash transfer programs on child nutritional status has been difficult to achieve due to the considerable logistic and ethical constraints that characterize humanitarian settings.

Methods

We present the findings from a quasi-experimental longitudinal study of a conditional emergency cash transfer program implemented by Concern Worldwide in 2012 during a food crisis in Tahoua, Niger, in which the use of a concurrent control group permits estimation of the program’s impact on child weight gain. Program beneficiaries received three transfers totaling approximately 65% of Niger’s gross national per capita income; mothers attended mandatory sessions on child and infant feeding and care practices. Dietary and anthropometric data from 211 vulnerable households and children targeted by the intervention were compared with 212 similarly vulnerable control households and children from the same 21 villages. We used multilevel mixed effects regression to estimate changes in weight and weight-for-height Z scores (WHZ) over time, and logistic regression to estimate the probability of acute malnutrition.

Results

We found the intervention to be associated with a 1.27 kg greater overall weight gain ( P < 0.001) and a 1.82 greater overall gain in WHZ ( P < 0.001). The odds of having acute malnutrition at the end of the intervention were 25 times higher among children in the comparison group than those in households receiving cash ( P < 0.001).

Conclusions

We conclude that this emergency cash transfer program promoted child weight gain and reduced the risk of acute malnutrition among children in the context of a food crisis. We suspect that the use of strategic conditional terms and a valuable transfer size were key features in achieving this result. Limitations in study design prevent us from attributing impact to particular aspects of the program, and preclude a precise estimation of impact. Future studies of this nature would benefit from pre-baseline measurements, more exhaustive data collection on household characteristics and transfer use, and further investigation into the use of conditional terms in emergency settings.