Tuesday, October 20, 2026

The AI Psych Brief

🔬 Section 1

Psychiatric Research Article

Individualized Connectivity-Guided Versus Conventional Targeting of Accelerated Theta-Burst Stimulation in Depression: A Randomized, Double-Blind, Parallel-Design Trial

Li M, Wu Z, Lu X, Zalesky A, Cheng X, Zhang Y, Cash RFH, et al.; Jiang T, Yan H · American Journal of Psychiatry · October 2026 (online August 12, 2026)

This single-center, double-blind trial from Peking University randomized 123 adults with major depression or bipolar II depression (at least one failed antidepressant) to an accelerated course of intermittent theta-burst stimulation — 20 sessions over 2 weeks with robotic neuronavigation — aimed at one of three left DLPFC targets. The comparison arms were the traditional 5-cm rule, a site chosen by each patient's resting-state functional anticorrelation with the subgenual anterior cingulate (sgACC), and a site chosen by structural (tractography) connectivity to the sgACC.

At week 2, structural-connectivity targeting beat the 5-cm rule by about 9 percentage points of HAM-D reduction (d = 0.70). By week 6 both connectivity-guided arms outperformed the scalp-measured target (structural d = 1.03; functional d = 0.75), but by week 12 the differences were no longer significant. Adverse events were similar across arms, with no seizures or switches into mania.

💡 Why it matters

Where the coil lands matters: individualized sgACC-anchored targeting looks like it speeds and deepens early TMS response, which is relevant as accelerated protocols spread. It is still one center, about 40 patients per arm, with an advantage that faded by three months — so treat it as strong preliminary evidence, not yet a reason to refer only to sites with MRI-guided targeting.

Read the paper → doi:10.1176/appi.ajp.20251084

📖 Section 2

Psychotherapy Research Article

A randomized controlled trial of regulation-focused psychotherapy for children with externalizing disorders

Halfon S, Albayrak M, Körlü Bulut M, Eğrilmez D, Gülzari E, Arikan K, Büyükaşik CJ, Hoffman L, Kufferath-Lin T, Prout TA, Rice T · Psychotherapy Research · September 30, 2026 (online)

Regulation-Focused Psychotherapy for Children (RFP-C) is a manualized, time-limited psychodynamic treatment for disruptive behavior. Rather than teaching behavioral rules, it works on the child's implicit avoidance of painful affect, treating oppositional behavior as a defense. Eighty Turkish children (mean age 8.5, 71% boys) with externalizing disorders were randomized to RFP-C or to an active comparator: a group combining parenting and child social-skills training.

At post-treatment, RFP-C reduced broad externalizing problems more than the comparator (d = -0.47). It also improved global functioning (d = 0.57) and lowered parenting stress. However, it did not separate on ODD symptoms, and the externalizing advantage shrank and lost significance by 6-month follow-up.

💡 Why it matters

This is one of the few randomized tests of a psychodynamic treatment for childhood disruptive behavior against an active, skills-based comparator, and the post-treatment signal is real. Read it as "promising, not yet durable." When parent-management training has stalled, an affect-regulation-focused dynamic approach is a defensible next step, but the gains may need booster work to hold.

Read the paper → doi:10.1080/10503307.2026.2738632

💊 Section 3

Psychiatric Fact

For lithium polyuria, amiloride blocks the door lithium uses to get into the cell

Lithium reaches the collecting-duct principal cell mainly through the apical epithelial sodium channel (ENaC). Once inside, it inhibits GSK-3β signaling and downregulates aquaporin-2, which causes nephrogenic diabetes insipidus. Amiloride blocks ENaC, cutting off lithium's entry at its source: in animal and cell work it lowered intracellular lithium in principal cells. In a placebo-controlled crossover in lithium-treated patients, 10 mg daily raised maximal urine osmolality and urinary AQP2.

That mechanism is why amiloride (typically 5–10 mg, up to 20 mg/day) is the preferred add-on when a patient who is doing well on lithium is up three times a night. A thiazide also reduces urine volume, but it does so by causing volume contraction, which increases proximal lithium reabsorption and pushes the level up. Amiloride is far less lithium-sparing than that, but it is not neutral: recheck the level and potassium after starting it, and avoid it with an ACE inhibitor or ARB unless you are monitoring closely. The evidence supports treating established polyuria; whether amiloride prevents it is unproven.

🛋️ Section 4

Psychotherapy Teaching Pearl

With a paranoid patient, make the interpretation about you, not them

John Steiner (1994) distinguished patient-centred interpretations ("you are pushing me away because…") from analyst-centred ones ("you experience me as someone who…"). A patient functioning in a paranoid-schizoid way, flooded and relying on projection, hears a patient-centred interpretation as the analyst handing the projected badness back. To that patient it is an accusation, so it confirms the persecutory transference rather than loosening it. An analyst-centred interpretation describes the patient's experience of the analyst — what the analyst is felt to be doing, wanting or feeling. It shows that the projection has been received and thought about, which is containment in Bion's sense, before the patient is asked to own anything. This is not a permanent stance. It is a sequencing decision: once the patient can tolerate being understood, patient-centred work becomes possible. Steiner also notes a trap: the analyst who stays with analyst-centred interpretations to avoid the patient's anger has stopped containing and is simply placating.

🗒️ Vignette

A 34-year-old woman with paranoid and narcissistic traits arrives saying the resident "obviously" rescheduled last week to get rid of her. The resident's first impulse is a patient-centred move: "You often conclude people are rejecting you when plans change." That is accurate, but in this state she will hear it as "the problem is you," and she will escalate or leave. Instead the resident says, "So right now I'm someone who moves you around to suit myself — and I'd probably deny it if you asked." She pauses. "Well — would you?" The resident does not reassure her or explain the scheduling. They say, "It sounds as though it would be hard to believe me either way." The heat drops. Several sessions later, after more of this, she brings up her father's unannounced moves on her own, and only then does the resident link the two.