Identifikasi Faktor Stunting pada Balita di Provinsi Riau Tahun 2023 dengan Pendekatan GLM dan INLA

Background: Stunting remains a chronic malnutrition issue in toddlers in Riau Province, Indonesia, with spatial disparities despite a low prevalence of 13.6% in 2023. Factors like low birth weight (BBLR), infant health services, active posyandu, maternal Fe3 tablet consumption, and toddler vitamin A coverage contribute to uneven cases across districts. Objective: To identify significant predictors of stunting cases and determine the best regression model using GLM (Poisson, Negative Binomial, Gaussian) and INLA. Method: This quantitative comparative study used secondary data from Riau Health Profile 2023 and BPS on 12 districts/cities (population: all Riau districts; sample: full dataset). Predictors were analyzed via VIF for multicollinearity, GLM with AIC/BIC, and INLA with DIC. Results: Negative Binomial GLM showed significant positive effect of BBLR (β=0.174, p=0.001) and negative effects of infant health services (β=-0.023, p=0.007) and active posyandu (β=-0.049, p=0.049); Fe3 and vitamin A were insignificant. Conclusion: Negative Binomial is the best-fit model (AIC=175.309, BIC=178.703), recommending targeted BBLR interventions for spatial equity.