Growth Monitoring and Promotion Model Quality, Regularity of Utilisation, and Child Nutritional Outcomes in High-Burden Communities: Evidence from a Convergent Parallel Mixed-Methods Study in Sulima Chiefdom, Falaba District, Sierra Leone

Background: Worldwide, an estimated 149 million children below the age of five are stunted and a further 45 million are wasted, with the heaviest concentration of this burden falling on Sub Saharan Africa. In Sierra Leone, where national stunting stands at 29% and climbs considerably higher across isolated northern districts, a pressing evidence gap remains that the existing crosssectional literature has not resolved using rigorous multivariable methods. Methods: We carried out a convergent parallel mixed-methods study made up of (1) a communitybased cross-sectional survey of 617 mother–child dyads and (2) twelve (12) key informant interviews (KIIs) examined through Framework Analysis in NVivo 14 (kappa = 0.82). Fieldwork took place in Sulima Chiefdom, Falaba District, between January and March 2026. Anthropometric indices were derived from the WHO 2006 Child Growth Standards using the WHO Anthro (igrowup) macro, after which the computed z-scores were transferred to SPSS for analysis. Independent correlates of regular GMP attendance (≥4 sessions/6 months) were established through bivariable and multivariable logistic regression, while multiple linear regression isolated the independent contribution of GMP attendance to height-for-age z-score (HAZ) and weight-for-age z-score (WAZ). A formal mediation analysis with bootstrapped inference (5,000 resamples) examined the Knowledge–Attitude–GMP Attendance pathway, and Spearman’s correlation was used to probe dose-response patterns. Results: Regular GMP attendance reached 41.8% (n = 258/617), a level 36 percentage points below ever-attendance (77.8%). Stunting prevalence stood at 28.7% (95% CI: 25.1–32.3%) and global acute malnutrition (GAM) at 18.3% (95% CI: 15.3–21.4%), surpassing the WHO emergency threshold. Children attending regularly showed markedly less stunting (19.0% vs. 35.7%; OR = 0.42, 95% CI: 0.29–0.62) and a mean HAZ that was 0.49 SD higher (Cohen’s d = 0.41, p < 0.001). Within the multivariable linear model, regular attendance held as an independent predictor of HAZ (β = 0.365, SE = 0.102, p < 0.001) and WAZ (β = 0.343, SE = 0.095, p < 0.001) once 11 covariates were taken into account. IYCF counselling (aOR = 4.20), positive attitude (aOR = 3.72), mother support group membership (aOR = 1.87), and perceived benefit (aOR = 1.79) each independently predicted regular attendance, whereas maternal knowledge lost significance entirely (aOR = 1.23, p = 0.807). Mediation analysis indicated that attitude carried an estimated 25.3% of the knowledge–attendance association (indirect effect a×b = 0.048; bootstrap 95% CI: 0.023–0.079). Dose-response patterns showed stunting falling steadily from 32.5% at zero sessions to 13.3% at six (ρ = 0.170, p < 0.001). The qualitative data supplied convergent mechanistic support for each quantitative result. Conclusion: GMP participation shows an independent and significant association with stronger child linear growth and ponderal status in one of Sierra Leone’s most underserved settings. The modifiable levers that matter are not individual knowledge but attitudinal, relational, and structural ones: quality service encounters that cultivate a positive attitude, IYCF counselling that fosters purposeful engagement, social capital generated through mother support groups, and physical accessibility. Taken together, the results argue that programmes should consider shifting away from attendance registration toward quality-centred integrated service delivery, reinforced by community mobilisation and satellite GMP infrastructure serving distant populations.