A noninvasive brain stimulation device shows a real, if modest, antidepressant signal
STUDY: Qin PP et al, JAMA Netw Open 2026;9(7)
STUDY TYPE: Randomized, double-blind, sham-controlled trial
FUNDING: Mental Health Research Centre (Hong Kong Polytechnic University), Strategic Hiring Scheme, Department of Rehabilitation Sciences (Hong Kong Polytechnic University)
Background
Most neuromodulation devices use magnets (TMS) or electrical stimulation (tDCS) to activate neurons. Transcranial pulse stimulation (TPS) relies on ultrasound. This noninvasive technique reaches structures deep in the brain, and early open-label work hinted at antidepressant effects. This is the first sham-controlled trial to test whether that holds up.
The Study
- 80 adults with major depressive disorder in Hong Kong, most already on stable antidepressants.
- Randomized to 12 sessions of active or sham stimulation targeting the left dorsolateral prefrontal cortex, three times a week for 4 weeks.
- Sham used an air-filled cushion that mimicked the sensation without delivering pulses.
- Primary outcome: change in depression scale (MADRS) score after treatment. Patients, raters, and operators were blinded.
TPS beat sham on the primary outcome, but the margin was thin. MADRS scores dropped 12.1 points with active treatment versus 7.9 with sham, a difference of about 4.7 points (effect size = 0.45, P = .048). That’s on the smaller end of what’s considered clinically meaningful.
Although the primary outcome was somewhat disappointing, response and remission rates were in line with what we see in TMS trials (response was 52.5% with active stimulation versus 25% with sham; remission was 40% versus 17.5%). Unlike TMS, TPS required only 12 sessions instead of 20.
On brain scans, TPS increased connectivity between the treatment target and several depression-related regions, including the orbitofrontal cortex and posterior cingulate cortex, areas known to be underconnected in depression. Oddly, these connectivity changes didn’t correlate with how much better patients felt.
Side Effects
Mild tingling and pain at the stimulation site were common (58% and 48% with active treatment versus 18% and 10% with sham), all resolving within 24 hours without treatment.
Limitations
Small trial, barely statistical significance (P = .048). Protocol adherence was only 66%, though nearly all participants still finished within 5 weeks. Some operators could tell active from sham stimulation based on feel and sound, a blinding weakness the authors acknowledge.
Practice Implications
- With one randomized trial, this isn’t ready for practice, but an encouraging sign worth watching.
—Chris Aiken, MD
Director, Psych Partners
Editor in Chief, Carlat Psychiatry Report
