The AI Psych Brief

Comparative effectiveness of mood stabiliser monotherapy versus second-generation antipsychotic monotherapy versus combination therapy for continuation and maintenance treatment after a first episode of psychotic mania in North America, Chile, and Spain: a target trial emulation using data from the BD-CAUSAL Collaboration

Szmulewicz AG, Liang YH, Martínez-Alés G, O'Dea R, et al.; Hernán MA, Öngür D; BD-CAUSAL Collaboration · The Lancet Psychiatry · October 2026

Young people with a first manic episode with psychotic features usually land in first-episode psychosis clinics, where second-generation antipsychotic (SGA) monotherapy is the default — yet there are almost no trials to guide maintenance after that first episode. Using linked first-episode cohorts from North America, Chile, and Spain, the authors emulated a pragmatic target trial comparing initiation of mood stabiliser monotherapy, SGA monotherapy, or the combination in 371 patients (median age 22).

Over two years, 42% were hospitalised for psychiatric reasons. The 2-year risk of psychiatric hospitalisation or ED visit was 42.3% with a mood stabiliser, 42.2% with combination therapy, and 49.5% with SGA monotherapy — a risk difference of about 7 percentage points (risk ratio 0.82) in favour of mood stabiliser-based strategies, with no difference between mood stabiliser alone and combination.

Read the paper →  doi:10.1016/S2215-0366(26)00239-7

Toward Personalized Psychotherapy in Anorexia Nervosa: Evidence from Long-Term Moderator Analyses of the ANTOP Study

Wild B, Friederich HC, Giel KE, Herzog W, et al.; Schmidt U, Zipfel S · Psychotherapy and Psychosomatics · August 2026

The ANTOP trial randomized 242 adult outpatients with anorexia nervosa at 10 German university hospitals to 10 months of focal psychodynamic therapy (FPT), enhanced CBT (CBT-E), or optimized usual care, and found no overall BMI difference between the arms. This exploratory secondary analysis asked a supervisor's question instead: for whom does each therapy work better? Using mixed models on the 160 patients randomized to FPT or CBT-E, the authors tested baseline moderators of BMI at the final follow-up.

Low depression severity and high mental quality of life at baseline favoured CBT-E, while high depression severity and low mental quality of life favoured FPT; the same direction held at every earlier timepoint. Across the broad middle of the distribution the two therapies performed similarly — the differences appeared at the extremes.

Read the paper →  doi:10.1159/000553469

Clonus, not temperature, is what separates serotonin toxicity from NMS

When a patient on a serotonergic drug and an antipsychotic becomes febrile and rigid, the neuromuscular exam discriminates better than the vitals. The Hunter criteria — built around clonus (spontaneous, inducible, or ocular) plus agitation, diaphoresis, tremor, or hyperreflexia in the setting of a serotonergic agent — were about 84% sensitive and 97% specific against toxicologist diagnosis, outperforming the older Sternbach criteria. Serotonin toxicity is hyperkinetic: clonus and hyperreflexia, worse in the legs, onset within hours of a dose change. NMS is bradykinetic: lead-pipe rigidity, reduced reflexes, evolving over days, with a markedly raised CK. The trap is severe serotonin toxicity, where truncal rigidity can mask clonus as temperature climbs above about 38.5°C — at that point management is sedation, cooling, and if needed paralysis with a non-depolarizing agent, since antipyretics don't work on muscle-generated heat. Getting it right changes treatment: cyproheptadine for one, and stopping all dopamine blockers (and avoiding antipsychotics for sedation) for the other.

In TFP, go to the channel that's carrying the most

Transference-focused psychotherapy describes three channels of communication: what the patient says, what the patient does (tone, posture, behaviour toward the frame), and what the therapist feels. With neurotic patients the verbal channel usually carries the dominant affect. With borderline organization it often doesn't — the words can be empty or contradictory while the real object relation is being enacted non-verbally or induced in the countertransference. The TFP tactic is to find the affectively dominant channel at this moment and work from there, even if the patient's stated topic seems more pressing. That means describing the dyad being lived out (for example, a contemptuous inspector and a failing subordinate) using nonverbal and countertransference data, rather than following the narrative. Ignoring the dominant channel is how sessions become fluent and useless.