Peer-reviewed article

Exploring outcomes between single and dual caregiver households participating in a parenting and family strengthening intervention in Rwanda: the case of Sugira Muryango

Joyeuse Ukwishaka, Vincent Sezibera, María Piñeros-Leaño, Theresa Stichick Betancourt · 25 Nov 2025

Abstract

INTRODUCTION: Early childhood is a critical period for development, yet families in low-resource settings often face barriers that compromise effective caregiving. Research evidence shows that parenting interventions can improve child growth and development. Given the unique challenges faced by single-headed households, this study examined whether caregiver household type modified the effects of Sugira Muryango, an evidence-based home-visiting program in Rwanda. Specifically, it assessed whether changes in child growth and development, dietary diversity, and healthcare-seeking behaviors over time differed between children from single versus dual caregiver households in the Sugira Muryango cluster randomized controlled trial (cRCT). METHODS: A total of 1,049 households were enrolled in the Sugira Muryango cRCT, with 508 households assigned to the control group and 541 to the intervention group. The analytic sample included 959 children from 959 households. Of these, 498 were from single caregiver households and 461 from dual caregiver households. Baseline and follow-up data collected 12 months post-intervention were analyzed. Linear mixed-effects models (LMMs) were used to analyze continuous outcomes, accounting for children nested within households and clusters. Models included fixed effects for timepoint, treatment group, caregiver household type (single vs. dual), and their three-way interaction (timepoint × treatment group × caregiver household type). Random intercepts for households and clusters were included to account for the hierarchical data structure. Mixed-effects logistic regression models were used for binary outcomes. RESULTS: There was no evidence of effect modification by caregiver household type on any outcome. For child growth indicators, three-way interaction terms were not significant for weight-for-age z-score (WAZ) (β = 0.081, 95% CI: -0.077 to 0.239), height-for-age z-score (HAZ) (β = 0.013, 95% CI: -0.180 to 0.207), or weight-for-height z-score (WHZ) (β = 0.129, 95% CI: -0.083 to 0.341). Dietary diversity also showed no differential effect by caregiver household type (β = 0.256, 95% CI: -0.175 to 0.686). For the Ages and Stages Questionnaire developmental domains, the three-way interaction terms were not significant: Communication (β = -0.272, 95% CI: -0.779 to 0.235), Gross Motor (β = 0.107, 95% CI: -0.571 to 0.785), Fine Motor (β = 0.129, 95% CI: -0.331 to 0.590), Problem-Solving (β = 0.101, 95% CI: -0.372 to 0.574), and Personal-Social (β = 0.090, 95% CI: -0.370 to 0.550). No significant modification effects were observed for care-seeking behaviors for diarrhea (β = 2.03, 95% CI: -0.92 to 4.97) or fever/cough (β = 1.22, 95% CI: -0.13 to 2.56). CONCLUSION: The Sugira Muryango intervention yielded similar outcomes across caregiver household types. Given its inclusive design and potential to deliver equitable benefits across diverse households settings in low-resource contexts, it offers a promising model for community-based interventions with important implications for scaling up similar programs.