Key Takeaways
Extended Takeaways
- Lurasidone separated from placebo on both anxiety domains over 6 weeks in bipolar I depression, with larger placebo-corrected differences for psychic anxiety than somatic anxiety in both monotherapy and adjunctive treatment.
- Baseline anxiety severity did not eliminate antidepressant benefit: in the monotherapy study, MADRS effect sizes were 0.53 in the lower anxiety group and 0.39 in the higher anxiety group, and CGI-BP-S score (Depression) effect sizes were 0.61 and 0.43, respectively.
- For patients receiving lurasidone monotherapy, baseline reduced sleep identified a subgroup with stronger anxiolytic signal, including HAM-A total Cohen d = 0.47 overall, 0.55 with 20–60 mg/d, and 0.38 with 80–120 mg/d.
- The lower monotherapy dose range showed the most consistent anxiety benefit when decreased sleep was present at baseline, with psychic anxiety Cohen d = 0.58 and somatic anxiety Cohen d = 0.35 versus 0.42 and 0.24, respectively, at 80–120 mg/d.
- Anxiety improvement tracked with functional change, not just mood change: HAM-A change was significantly associated with SDS improvement in pooled lurasidone treatment, and in monotherapy the HAM-A–SDS mediation remained significant after accounting for MADRS improvement (P = .044, F1,395 = 4.08).
- Because placebo response rose as baseline HAM-A thresholds increased, especially for depressive symptoms, clinicians should be cautious about assuming that more anxious bipolar depression predicts less antidepressant response to lurasidone.
