The Relational Cascade Model: a Measurement for Change framework for testing caregiver coaching as upstream mental health promotion
Frontiers in Public Health · 2026-11-04 · 14
DOI: 10.3389/fpubh.2026.1943530 ↗Rezumat
Early childhood development remains chronically underfunded and marginalized in global health policy, often treated as a downstream education concern because no single sector owns it. We propose that the relational environment of early childhood, mediated through caregiver regulation, attunement, and responsiveness, is a proximal pathway through which adversity may be transmitted across generations and therefore interrupted. The Relational Cascade Model (RCM) is an architecture linking trainer, coach, caregiver, and child in a cascade of regulatory capacity in which each level requires the same relational resources it transmits. The RCM draws on developmental cascade theory, attachment and co-regulation, salutogenesis, differential susceptibility, and neurobiological evidence on hippocampal plasticity, and is operationalized through the Conscious Discipline curriculum, which is distinct from the cascade itself. We describe its implementation in two low- and middle-income country contexts, via the Roving Caregiver Programme in Grenada and Community Health Promoter coaching in Kenya, reaching 5,692 children through 2,828 trained parents and 279 community workers as of June 2026. The model is assessed through a Measurement for Change (M4C) framework capturing caregiver wellbeing, intervention fidelity, child neurodevelopment, and attitudes toward corporal punishment. These are captured within one cross-sectoral architecture, making the caregiving relationship visible across health, education, and child-rights accountability streams. We argue that this measurement framework, rather than additional intervention content, supplies the multilevel data needed to test the cascade rather than assume it, and propose it as a template for moving early childhood, and public health broadly, from siloed accountability to systems-level learning.
EUVIMED