From Prediction to Adaptation: Rethinking the Epistemic Role of Inhalation Toxicology

Abstract

Inhalation toxicology has long aimed to predict the health effects of airborne substances before exposure occurs, relying on stable dose-response relationships and well-characterized hazards. This approach becomes increasingly limited when confronted with emerging materials, complex mixtures, and dynamic exposure scenarios, where key mechanisms and variables are not fully known in advance. In this Perspective, we propose reframing inhalation toxicology from a predictive toward an adaptive science, in which experimental and computational systems are designed to rapidly generate and integrate information under conditions of uncertainty. We outline how flexible in vitro exposure models, computational dosimetry, and iterative evidence integration can form adaptive frameworks that support learning and updating rather than static prediction. We further discuss the implications of this shift for experimental design, model evaluation, and the handling of uncertainty. This conceptual reframing offers a scientifically grounded approach for maintaining relevance and rigor in inhalation toxicology as exposure landscapes continue to evolve.