Study compares DSM-5 depressive symptom profiles of self-reported ADHD and ASD Twitter users

This study analyzed 1,282,437 tweets from 792 users with self-reported ADHD or ASD diagnoses to examine population-level differences in how they express DSM-5 depressive symptoms. Using MentalRoBERTa fine-tuned on ReDSM5, the researchers classified tweets into nine symptom categories and applied L1-penalised logistic regression to distinguish between the groups.

• MentalRoBERTa achieved a macro-F1 of 0.901 on a held-out set, outperforming the original ReDSM5 benchmark.
• ADHD vs ASD classification yielded stable but modest performance with a cross-validated ROC-AUC of 0.645-0.653.
• Cognitive issues, sleep issues, appetite change, and fatigue leaned toward ADHD, while suicidal ideation and anhedonia leaned toward ASD.
• A largely shared symptom co-occurrence structure emerged between groups, with no pair meeting the criterion for a robust disorder-specific difference.

The findings indicate that population-level differences in depression-related language are consistently observed across filtering thresholds, reflecting reproducibility rather than clinical validity. The authors conclude that these results are exploratory and do not establish differing phenomenology at the individual level.