The AI Psych Brief

A Phase 2 Randomised, Double-Blind, Placebo-Controlled Trial of Emestedastat in Patients With Major Depressive Disorder and Cognitive Impairment (XanaCIDD Trial)

Taylor JD, Rolan PE, Jaros MJ, et al. · British Journal of Psychiatry · July 2026

165 patients with persistent MDD and objectively confirmed cognitive impairment (most on background antidepressants) were randomized to emestedastat 10 mg, a brain-penetrant inhibitor of intracellular cortisol synthesis, or placebo for 6 weeks with 4 weeks of blinded follow-up. The primary cognitive endpoint showed no benefit over a large placebo response, but a prespecified secondary depression endpoint (MADRS) showed a modest, delayed benefit peaking at week 10 (2.7-point difference, Cohen's d 0.43); about 10% of drug-treated patients had mild-moderate transaminase elevation.

Read the paper →  doi:10.1192/bjp.2026.10689

Cognitive Mechanisms of Change in OCD: Inferential Confusion, Obsessive Beliefs, and Mindfulness Across Three Randomized Treatments

Ouellet-Courtois C, Bouchard S, Giguère CÉ, et al. · Behaviour Research and Therapy · June 2026

This mechanism-focused reanalysis of a 111-patient multicenter trial compared CBT-with-ERP, inference-based CBT, and mindfulness-based stress reduction for OCD, testing which candidate mechanism — inferential confusion, obsessive beliefs, or dispositional mindfulness — actually drove improvement. All three treatments reduced inferential confusion and obsessive beliefs, but mindfulness itself didn't change significantly in any arm. When the mechanisms were pitted against each other in the same model, only reduction in inferential confusion — mistaking imagined possibilities for reality — independently predicted symptom improvement, across all three treatments.

Read the paper →  doi:10.1016/j.brat.2026.105022

The Lorazepam Challenge Has a Ceiling Most Residents Don't Know About

The catatonia challenge is 1–2 mg IV lorazepam, reassessed at 5 minutes (15 minutes IM, 30 minutes oral) — and the ceiling worth remembering is that 4 mg buys no additional diagnostic yield over 2 mg. A positive response is a roughly 50% reduction in signs on a standardized scale like the Bush-Francis, usually visible within 10 minutes. Some signs respond to either dose — stupor, mutism, staring, posturing, withdrawal — while others, like echolalia, rigidity, negativism, and mitgehen, respond selectively at the higher end, so a partial response with residual negativism after 2 mg isn't necessarily a failed test. The practical error is repeatedly escalating the dose chasing a cleaner response; if 2 mg produces no meaningful change, more lorazepam won't clarify the picture — it's a cue to reconsider the diagnosis or move toward ECT rather than dose-stack.

Interpersonal Deficits Treats Isolation as the Illness, Not Its Symptom

Of IPT's four problem areas, interpersonal deficits is the one residents most often misuse — treating chronic social impoverishment as background context for the depression rather than as the treatment target itself. The other three areas, grief, role dispute, and role transition, all locate the problem inside a specific relationship or change; deficits is reserved for patients without the relational material to generate one, and technique shifts accordingly. Instead of working an existing relationship, the therapist works the therapeutic relationship itself as the only available interpersonal data, using it to build a first new connection in real time. Treating a deficits presentation with dispute or transition technique fails because there's no target relationship to negotiate change within — the absence of one is the whole problem.