Abstract
BACKGROUND: Anxiety and depression are highly prevalent among pre-university students, often intensified by the academic stress associated with entrance exams. Linguistic analysis of written texts offers a promising, non-invasive approach for early detection and prevention. Study design: Cross-sectional study. OBJECTIVES: To examine the association between linguistic features in essays and levels of anxiety and depression, identifying specific language patterns linked to these conditions. SETTING: Pre-university preparatory courses in São Paulo, Brazil, in 2023. METHODS: Participants were 62 pre-university students (51 females, 11 males; M = 20.3 years, SD = 2.65) who completed a self-report form shared via WhatsApp or in-person at preparatory schools. The form included sociodemographic questions, the GAD-7 and PHQ-9 scales, and the upload of an argumentative essay written within the previous month as part of their regular coursework. Essays were analyzed using LIWC software, and multivariate regression models identified linguistic features associated with anxiety and depression scores. RESULTS: Higher anxiety levels correlated with increased use of words related to affiliation and home, and decreased use of leisure and money-related terms. Depression was associated with higher frequency of drives and number-related words, and fewer motion-related terms. CONCLUSION: Linguistic analysis can assist in identifying emotional distress among pre-university students, offering a potential tool for early screening and intervention in educational and mental health contexts.