A single-centre randomised, double-blind, placebo-controlled crossover trial of intranasal oxytocin in behavioural variant frontotemporal dementia found no improvement on its primary outcomes, the facial emotion and intention recognition tasks, and reported reduced accuracy in recognising harmful intentions, while exploratory measures pointed the other way. The trial, published on 22 January 2026 in Frontiers in Aging Neuroscience by Valentina Colonnello, Katia Mattarozzi and colleagues at the University of Bologna and the IRCCS Istituto delle Scienze Neurologiche di Bologna, enrolled 20 patients recruited between 2019 and 2022. Each received 24 IU of oxytocin (Syntocinon spray, six puffs, three per nostril) administered by a nurse on two consecutive days, then, after a one-week washout, placebo for two days, in randomised order. Sprays were prepared and blinded by the pharmacy of Heidelberg University Hospital.

On the secondary and exploratory measures, the picture was different. Oxytocin was safe and well tolerated, and carers reported changes in everyday social behaviour including greater awareness, more spontaneous initiative, more socio-affective engagement and reductions in appetite and impulsivity. Exploratory analyses, which the authors describe as exploratory and not as confirmatory evidence, found patients were less likely to misclassify fearful and sad faces as angry, and used fewer first-person pronouns during a semi-structured interview analysed with the Linguistic Inquiry and Word Count method. The authors conclude that oxytocin appeared to reduce threat bias, promote social engagement and decrease self-focused language, while nothing improved on the experimental tests of social cognition.

The result lands in a field with no approved treatment for the behavioural symptoms of frontotemporal dementia, symptoms that drive much of the distress reported by families. It follows the multicentre FOXY trial, a phase 2a/2b adaptive crossover study of 72 IU intranasal oxytocin for apathy in frontotemporal dementia across 11 clinics in Canada and the United States, published in The Lancet Neurology in 2025, which found dosing every third day was well tolerated and associated with a small reduction in apathy on the Neuropsychiatric Inventory. An earlier randomised dose-finding study led by Elizabeth Finger, published in Neurology in 2015, established short-term safety and tolerability at 24, 48 and 72 IU twice daily for a week in 23 patients.

The Bologna authors themselves set out the counterpoint. The final sample of 20 is underpowered for medium-sized effects, some results sat close to the conventional significance threshold, there was no long-term follow-up, and expectancy effects cannot be excluded despite blinding. The reduced accuracy in recognising harmful intentions is reported as a finding, not a benefit, and the authors call for replication in larger samples before conclusions are drawn about what oxytocin does to higher-order social processing in this disease.