Background

Cash-based interventions (CBIs), offer an interesting opportunity to prevent increases in wasting in humanitarian aid settings. However, questions remain as to the impact of CBIs on nutritional status and, therefore, how to incorporate them into emergency programmes to maximise their success in terms of improved nutritional outcomes. This study evaluated the effects of three different CBI modalities on nutritional outcomes in children under 5 y of age at 6 mo and at 1 y.

Methods and findings

We conducted a four-arm parallel longitudinal cluster randomised controlled trial in 114 villages in Dadu District, Pakistan. The study included poor and very poor households (n = 2,496) with one or more children aged 6–48 mo (n = 3,584) at baseline. All four arms had equal access to an Action Against Hunger–supported programme. The three intervention arms were as follows: standard cash (SC), a cash transfer of 1,500 Pakistani rupees (PKR) (approximately US$14; 1 PKR = US$0.009543); double cash (DC), a cash transfer of 3,000 PKR; or a fresh food voucher (FFV) of 1,500 PKR; the cash or voucher amount was given every month over six consecutive months. The control group (CG) received no specific cash-related interventions. The median total household income for the study sample was 8,075 PKR (approximately US$77) at baseline. We hypothesized that, compared to the CG in each case, FFVs would be more effective than SC, and that DC would be more effective than SC—both at 6 mo and at 1 y—for reducing the risk of child wasting. Primary outcomes of interest were prevalence of being wasted (weight-for-height z-score [WHZ] < −2) and mean WHZ at 6 mo and at 1 y.

The odds of a child being wasted were significantly lower in the DC arm after 6 mo (odds ratio [OR] = 0.52; 95% CI 0.29, 0.92; p = 0.02) compared to the CG. Mean WHZ significantly improved in both the FFV and DC arms at 6 mo (FFV: z-score = 0.16; 95% CI 0.05, 0.26; p = 0.004; DC: z-score = 0.11; 95% CI 0.00, 0.21; p = 0.05) compared to the CG. Significant differences on the primary outcome were seen only at 6 mo. All three intervention groups showed similar significantly lower odds of being stunted (height-for-age z-score [HAZ] < −2) at 6 mo (DC: OR = 0.39; 95% CI 0.24, 0.64; p < 0.001; FFV: OR = 0.41; 95% CI 0.25, 0.67; p < 0.001; SC: OR = 0.36; 95% CI 0.22, 0.59; p < 0.001) and at 1 y (DC: OR = 0.53; 95% CI 0.35, 0.82; p = 0.004; FFV: OR = 0.48; 95% CI 0.31, 0.73; p = 0.001; SC: OR = 0.54; 95% CI 0.36, 0.81; p = 0.003) compared to the CG. Significant improvements in height-for-age outcomes were also seen for severe stunting (HAZ < −3) and mean HAZ. An unintended outcome was observed in the FFV arm: a negative intervention effect on mean haemoglobin (Hb) status (−2.6 g/l; 95% CI −4.5, −0.8; p = 0.005). Limitations of this study included the inability to mask participants or data collectors to the different interventions, the potentially restrictive nature of the FFVs, not being able to measure a threshold effect for the two different cash amounts or compare the different quantities of food consumed, and data collection challenges given the difficult environment in which this study was set.

Conclusions

In this setting, the amount of cash given was important. The larger cash transfer had the greatest effect on wasting, but only at 6 mo. Impacts at both 6 mo and at 1 y were seen for height-based growth variables regardless of the intervention modality, indicating a trend toward nutrition resilience. Purchasing restrictions applied to food-based voucher transfers could have unintended effects, and their use needs to be carefully planned to avoid this.