Judgment behaviors differentiate depression from anxiety

Depression and anxiety are common mental health disorders that often share overlapping symptoms. This can limit the specificity of standard screening tools and complicate differential diagnosis. Changes in judgment can present for both conditions and can be mathematically extracted from a short picture rating task. The goal of this study was to determine if these judgment features can differentiate anxiety from depression. This cross-sectional population study surveyed 3476 adult participants (ages 18-70) from across the United States in December 2021. Anxiety (State Trait Anxiety Inventory (STAI)) and depression (Patient Health Questionnaire 9 (PHQ-9)) scores categorized as moderate-to-severe were classified using random forest, with judgment variables and demographics used as predictors. Judgment variable differences were assessed across anxiety and depression symptom severities. The study cohort was predominately middle aged (mean 51), female (61%), white (84%), married (51%), with 5% self-reporting severe depressive symptoms and 28% reporting severe anxiety symptoms. The 15 judgment variables, with and without demographic variables, produced AUROCs of 0.798 and 0.793, respectively. Judgment variables, collectively, carried 83% of the predictive feature importance, whereas the seven demographics, apart from age, were the least important features (17% collectively). Judgment variables further differed across PHQ-9 and STAI severities. Judgment features from a picture rating task differentiated anxiety from depression and varied meaningfully across symptom severities, reflecting motivational deficits and emotional blunting. Given the prevalence of depression and anxiety, a quick, low-cost, and unbiased tool could improve diagnostic efficacy and patient well-being.