Modeling Coupled Wasting-Stunting Dynamics and Chronic Undernutrition Persistence in Indonesia

Abstract

Child undernutrition remains a major global public health challenge, with wasting and stunting representing acute and chronic manifestations of nutritional failure, respectively. Although strong epidemiological associations between wasting and stunting have been widely reported, the long-term dynamical mechanisms governing their interaction at the population level remain insufficiently understood. This study develops a coupled wasting--stunting dynamical model to investigate the interaction between acute and chronic undernutrition among children under five years of age in Indonesia. The proposed framework integrates wasting progression, severe wasting dynamics, chronic stunting accumulation, and nonlinear interaction mechanisms linking acute nutritional deterioration and chronic growth failure. The model additionally incorporates low birth weight recruitment as an early-life nutritional vulnerability mechanism. Model parameters were obtained through demographic assignment, literature-informed specification, empirical approximation, and semi-mechanistic calibration using Indonesian national nutritional surveillance data covering the period 2013--2024. The analytical results revealed a hierarchical dynamical structure governed by the wasting progression threshold $\mathcal{R}_W$. The analysis demonstrated that persistent wasting exposure may amplify long-term chronic undernutrition through nonlinear interaction pathways, while chronic growth failure simultaneously increases susceptibility to recurrent wasting progression. Local stability analysis showed that the wasting subsystem remains asymptotically stable under the calibrated parameter configuration satisfying $\mathcal{R}_W<1$. Model robustness and parameter influence were further evaluated using profile likelihood analysis, bootstrap-based uncertainty assessment, local sensitivity analysis, and variance-based global sensitivity analysis using Sobol indices. The sensitivity analyses demonstrated that wasting dynamics are governed primarily by deterioration and recovery processes, whereas long-term stunting persistence is dominated mainly by baseline chronic progression mechanisms. Long-term projections under multiple intervention scenarios showed that recovery-oriented interventions substantially reduce wasting and severe wasting prevalence, but produce only relatively modest reductions in stunting prevalence over the period 2024--2045. Even under intensive intervention scenarios, the projected stunting prevalence remained above the national target threshold of 14%, indicating the presence of strong structural persistence within the chronic undernutrition subsystem. Overall, the study suggests that chronic undernutrition should not be interpreted merely as an isolated nutritional condition, but rather as an emergent population-level consequence of sustained nutritional exposure, nonlinear interaction dynamics, and long-term feedback accumulation. The proposed framework provides a quantitative basis for evaluating long-term nutritional intervention strategies in settings with limited longitudinal nutritional transition data.