Generalization of respiratory symptom triggers

Abstract

Behavioral management of asthma and other chronic conditions depends upon the accurate identification of environmental factors that trigger symptom onset. In this study, we developed a lab-based conditioning method to study category-based acquisition and generalization of respiratory symptom triggers. During trigger acquisition, unique exemplars of two different categories were shown to a sample of healthy participants (n=48). CS+ exemplars were paired with CO2 inhalation on 50% of trials, while CS- exemplars were always paired with room air. Trigger categories differed in their conceptual similarity. In a generalization task, participants rated symptom expectancy for a set of triggers that included previously seen exemplars, novel exemplars, and exemplars from novel categories. Results show that participants acquired differential symptom expectancies based on category information, which generalized to novel CS+ exemplars and novel categories that shared similarity with the CS+ category. Greater similarity between CS+ and CS- categories increased differential effects for both old and novel exemplars of CS+ and CS- categories, and increased the proportion of novel CS+ exemplars that were remembered as being seen during acquisition. These findings suggest that a more narrowly defined contrast between triggers and nontriggers promotes category-based inference and could help to reduce uncertainty about potential triggers.