Opparounds

Pharmacological interventions for ADHD: a systematic review and dose-effect network meta-analysis

Nourredine M, Jurek L, Hamza T, Cipriani A, Subtil F, Parlatini V, Farhat LC, Veronesi GF, Efthimiou O, Salanti G, Cortese S · The Lancet Psychiatry · June 2026

This systematic review and dose-response network meta-analysis pooled 113 double-blind randomized trials — 14,138 children and adolescents, 11,016 adults — to map how ADHD medication efficacy actually changes across the licensed dose range, using a Bayesian model with restricted cubic splines to combine direct and indirect trial evidence.

In youth, efficacy for methylphenidate, amphetamines, and guanfacine rose steadily up to roughly 45 mg/day, 25 mg/day, and 4 mg/day respectively, with essentially no added benefit above those points. In adults, amphetamines plateaued near 50 mg/day, while methylphenidate kept climbing toward the same ceiling but with sharply diminishing returns per extra milligram. Across both age groups, pushing doses past the plateau — including above licensed maximums — bought little additional symptom control while increasing side-effect burden.

Read the paper →  doi:10.1016/S2215-0366(26)00091-X

Assessment of adverse effects attributed to statin therapy in product labels: a meta-analysis of double-blind randomised controlled trials

Reith C, Blackwell L, Emberson JR, Preiss D, Spata E, Davies K, et al., for the Cholesterol Treatment Trialists' Collaboration · The Lancet · February 2026

This individual-participant meta-analysis pooled more than 150,000 people across 23 double-blind trials — 19 placebo-controlled, 4 comparing higher- versus lower-intensity statin therapy — to test whether the 66 adverse effects listed on statin product labels actually occur more often on drug than on comparator. Blinding is the point: open-label and observational statin studies are notoriously contaminated by nocebo reporting once a patient knows what pill they're taking.

Memory impairment, depression, sleep disturbance, and sexual dysfunction each occurred at identical annual rates (about 0.2%) in the statin and comparator arms. After correcting for multiple comparisons, only 4 of the 66 labeled effects — deranged liver enzymes, other LFT abnormalities, altered urinary composition, and edema — showed a real, small excess risk with statins.

Read the paper →  doi:10.1016/S0140-6736(25)01578-8

Clozapine's other lethal side effect has nothing to do with the ANC

Clozapine's strong antimuscarinic and 5-HT3-antagonist activity slows gut transit more than any other antipsychotic, and clozapine-induced gastrointestinal hypomotility (CIGH) — constipation progressing silently to ileus, bowel ischemia, or perforation — kills roughly as many clozapine patients as agranulocytosis does, yet gets a fraction of the monitoring attention. Unlike neutropenia, there is no lab value to trend: by the time a patient reports obstipation, the bowel may already be compromised, because clozapine also blunts the visceral pain that would normally announce it. The practical fix is prophylactic, not reactive — a standing stimulant or osmotic laxative at initiation, a structured bowel-habit question at every visit rather than "any GI issues?", and treating any other anticholinergic or opioid on the regimen as a multiplier of risk. Constipation lasting more than three days, or any vomiting, distension, or reduced bowel sounds, deserves the same urgency as a fever in a clozapine patient — it can be just as fatal, just slower to announce itself.

Holding is a technique, not the absence of one

Winnicott's holding environment is often flattened into "just be warm and consistent," but it is a specific technical choice: an active decision to withhold interpretation when the patient's ego cannot yet use one. When a patient is flooded, fragmenting, or dissociating, offering content — even an accurate reading of the defense — asks an ego that cannot currently think to do more thinking, and it re-enacts the original failure of containment. Holding substitutes the therapist's steady presence and management of the frame — pace, tone, session boundaries — for interpretive content until the ego reconstitutes enough to use it. The skill is diagnostic as much as technical: recognizing the moment interpretation stops being facilitative and starts being intrusive, and knowing that returning to interpretation too early is itself a technical error, not extra effort.