Linguistic markers of emotional reactivity and their association with anxiety, depression, and stress among emergency call takers and dispatchers

Abstract

Emergency call takers and dispatchers (ECDs) are routinely exposed to emotionally intense and highly stressful events, placing them at elevated risk for mental health challenges. Despite their critical role in emergency response, ECDs remain underrepresented in occupational mental health research, and it is unclear why some adapt more effectively than others under chronic stress. Cognitive appraisal theory suggests that differences in emotional reactivity—how individuals evaluate stimuli in terms of arousal and valence—may contribute to this variation. To test this, a sample of ECDs provided open-ended narratives of a stressful work-related event and completed the Depression, Anxiety, and Stress Scale (DASS-21). Narratives were analyzed using natural language processing to quantify emotional reactivity through the usage of high and low arousal language, and positively and negatively valenced language. Analyses revealed that greater use of negatively valenced language significantly predicted higher depression and anxiety, whereas positively valenced and arousal language did not. Stress symptoms were not significantly predicted by any linguistic markers of emotional reactivity. Follow-up analyses showed that individuals with moderate or higher depression and anxiety used significantly more negatively valenced language than those below threshold. These findings suggest that negatively valenced language serves as a robust marker of psychological distress among ECDs, even when narratives are elicited in contexts that naturally evoke negative emotion. Results highlight the potential of language-based tools for early, unobtrusive detection of distress, offering scalable approaches to promote resilience and support well-being in this critical workforce.