This study uses a mixed-methods design. First, we conduct a quasi-experiment, enrolling ~600 mother-child pairs (300 in villages close to newly established PAs and 300 in similar sites without a PA), with surveys done in the year of PA implementation and two years later. Quantitative data will include demographic and socio-economic characteristics, dietary behaviour, and child growth measurements. We will use difference-in-differences method to estimate the intervention effects. Second, qualitative data will be collected through Focus Group Discussions (FGDs) and applying a participatory impact mapping approach. We will explore community views on how PAs impact the pathways to change, including access to resources or income. We will synthesize quantitative and qualitative data to identify the characteristics of an optimal PA for nutrition and health benefits at the Horn of Africa. Third, the amplification potential of this optimal PA will be identified through in-depth interviews with key informants, including transfer to other regions and sustaining the success of projects.