High-quality early childhood education can fundamentally alter children’s long-term education, earnings and well-being. In low-resourced settings, children’s development is hampered by undernutrition, poverty and limited access to services. This study will generate high-quality evidence on child development from children who both attend and do not attend daycare across informal settlements in Kenya by (1) examining the relationship between child development, nutritional status and household characteristics using a large, community-based sample and (2) measuring the impact of exposure to improvements in childcare quality on child development and nutritional status.

We combine a cross-sectional observational study of child development and anthropometrics among approximately 4700 children aged 0–5 across 11 counties in Kenya with a cluster randomised controlled trial that measures the impact of improved daycare quality on these outcomes. We use International Development and Early Learning Assessment to measure child development and record child height and weight to measure height-for-age and weight-for-height z-scores. Primary analyses will (1) estimate associations between child development outcomes and nutritional status using multivariable regression models, adjusting for prespecified covariates; and (2) examine differences in outcomes between children attending daycares in communities exposed to quality improvements and children attending daycares in control communities. Subgroup analyses will examine heterogeneity by child sex and daycare participation.

The Strathmore University Institutional Scientific and Ethical Review Committee (SU-ISERC) has provided ethical review for this study, with initial approval SU-ISERC1602/23. This study has also received IRB approval from The Ohio State University (#2023B0300). Written informed consent will be obtained from caregivers before participation. We will disseminate findings through peer-reviewed publications, policy briefs and presentations to local stakeholders, and we will publish de-identified data and replication code on a public repository.

AEARCTR-0011747.