Friday, August 7, 2026

Opparounds

🔬Psychiatric Research Article

Circadian Rhythm Stabilization App to Prevent Mood Episode Recurrence in Patients With Mood Disorders: A Multicenter, Double-Blind, Sham-Controlled, Randomized Clinical Trial

Yeom JW, Jeong J, Moon E, Park YM, Lee MS, Yoon HK, Cho CH, Lee HJ, et al. · American Journal of Psychiatry · June 2026

Korean investigators randomized 93 adults with major depressive disorder or bipolar disorder to twelve months of an active circadian-rhythm app or a sham app with an identical interface, in a genuinely double-blind design — the sham returned nonactionable feedback from a dummy algorithm. The active app used passive smartphone sensor data to generate individualized three-day mood forecasts and machine-learning-derived feedback aimed at stabilizing sleep-wake and activity rhythms.

In the modified intention-to-treat sample of 80 patients, the sham group had a substantially higher recurrence rate (incidence rate ratio 3.39, 95% CI 1.86 to 6.17), more cumulative episode-days per person-year (rate ratio 2.76, 95% CI 1.19 to 6.40), and shorter time to recurrence (hazard ratio 3.03, 95% CI 1.58 to 5.81). No significant adverse effects were reported. The sample is small and single-country, and the confidence intervals are correspondingly wide.

💡 Why it matters

This is one of the few digital mental health trials with a credible sham control, and the effect it found is on the outcome that actually matters in mood disorders — recurrence, not symptom scores. Treat it as strong support for the old chronotherapeutic instruction you already give (regularize sleep, wake, and activity times), now with a randomized number attached, rather than as an endorsement of any particular app.

Read the paper → doi:10.1176/appi.ajp.20251008

🩺General Medicine Article

Prolonged Short Sleep and Its Effect on Body Weight and Composition: A Pooled Analysis of Randomized Trials

Zuraikat FM, Scaccia SE, Cochran JA, Cheng B, Diaz KM, Aggarwal B, Jelic S, St-Onge MP, et al. · Annals of Internal Medicine · July 2026

Two randomized crossover trials at Columbia were pooled to ask whether *mild* chronic sleep loss actually causes weight gain, rather than merely traveling with it. Ninety-five adults at elevated cardiometabolic risk who habitually slept seven or more hours completed six weeks of adequate sleep and six weeks of sleep restricted by 1.5 hours per night, in randomized order with a multiweek washout — an unusually long exposure for a sleep experiment, and one that models real life better than a night of total deprivation.

Actual sleep fell by 78 minutes per night. Restriction produced small but consistent increases in body weight (0.45 kg), waist circumference (0.52 cm), and whole-body volume, along with higher leptin and 17 more sedentary minutes per day. The effects are modest and six weeks is too short to move body composition, so this establishes direction and plausibility rather than magnitude.

💡 Why it matters

Nearly every lever we pull in psychiatry touches sleep duration — the mania that shortens it, the insomnia that fragments it, the shift work that truncates it, and the sedating drug that appears to fix it — and this is randomized evidence that losing 90 minutes a night is itself weight-active, partly through moving less rather than eating more. In a patient starting an antipsychotic, sleep timing is a metabolic intervention you can begin the same day you order the baseline A1c and lipids.

Read the paper → doi:10.7326/ANNALS-25-01660

💊Psychiatric Fact

The dose-response curve for antipsychotics flattens far below where most services stop raising the dose

Leucht's dose-response meta-analysis of acute-phase trials put the ED95 of olanzapine near 15 mg/day, risperidone near 6.3 mg/day, and haloperidol near 6.3 mg/day — the dose delivering 95% of the achievable symptom reduction. Above those points the curve is essentially flat for positive symptoms, and for several agents, including haloperidol and aripiprazole, it is frankly bell-shaped: efficacy for negative symptoms declines as the dose climbs. So the reflexive escalation to haloperidol 20 mg or olanzapine 40 mg in an agitated inpatient is not buying antipsychotic effect. It is buying dopamine blockade above the therapeutic window, which is a different drug clinically — akathisia, parkinsonism, secondary negative symptoms, and a patient who becomes harder to distinguish from a deteriorating illness.

The corollary is what you do instead when a patient is not improving. Early response predicts eventual response: a patient who has not achieved roughly 20% symptom improvement by two weeks on an adequate dose is unlikely to respond to that agent by week six, which makes two weeks — not six or eight — the moment to switch drugs or reconsider the diagnosis, adherence, and substance use. In practice the two errors compound: the dose gets raised at week two because it feels like action, the side effects arrive, adherence falls, and the trial that should have ended at two weeks is still running at week ten. Change the drug, not the number.

🛋️Psychotherapy Teaching Pearl

In IPT, the problem area you choose is the treatment — pick the wrong one and the technique stops working

Interpersonal psychotherapy is often taught as if the interpersonal inventory were history-taking and the four problem areas — grief, role dispute, role transition, interpersonal deficits — were labels applied afterward. They are not labels; each one prescribes a different set of moves, and choosing between them is the single highest-leverage decision in the treatment. A role dispute sends you toward the stage of the dispute (renegotiation, impasse, dissolution), toward communication analysis, and toward what each party actually wants. A role transition sends you somewhere quite different: you help the patient mourn the old role, in detail and with its losses named, before you build competence in the new one.

The common technical error is formulating a transition as a dispute, because a transition almost always generates conflict and the conflict is the loud thing in the room. Work it as a dispute and you spend the treatment coaching the patient to negotiate with a spouse or a boss over what is really an unmourned identity — the sessions feel busy and the depression does not move. The diagnostic question is simple to ask and easy to skip: has something about who this person is required them to change, or are two people with intact roles pulling in different directions? IPT is time-limited, so a formulation error is not merely inelegant; it consumes the treatment.

🗒️ Vignette

A 52-year-old surgical nurse is referred with a four-month depressive episode after being moved off nights into a daytime administrative role following a back injury. She presents it as a fight: her husband thinks she should be grateful, her manager keeps changing her duties, and she is furious with both. The interpersonal inventory fills up with grievances, and the obvious formulation is a role dispute with the husband.

The resident's supervisor asks one question that redirects the case — what does she call herself now when someone asks what she does? She says, after a long pause, that she says she "used to be" a nurse. That is the formulation: a role transition dressed as a dispute. The fight with the husband is the transition's exhaust, not its engine, and negotiating it would have left the actual loss untouched.

So the work changes shape. Sessions three through six are spent on the old role in specific, sensory detail — the competence of running a difficult night, being the person the residents called, the physical language of the OR — and on what she does not miss, which she has never been permitted to say out loud (the exhaustion, missing her daughter's childhood, dreading the pager). Only then does the treatment turn to the new role: what a person with thirty years of clinical judgment can actually do in an administrative seat, and which relationships need to be built there. The marital conflict subsides without ever having been the target, because she stops needing her husband to grieve the job for her.