A type of transcranial magnetic stimulation could improve social communication in children with autism, according to a clinical trial

This press release is broadly accurate and reflects a well-conducted, relatively large randomised controlled trial. The findings are encouraging and suggest that accelerated continuous theta burst stimulation, a rapid form of patterned magnetic brain stimulation delivered using TMS equipment, may improve social communication in a group of autistic children for one month after treatment.

That said, the results should be interpreted with some caution. In this trial, the active stimulation group had a higher average SRS-2 score than the sham group at baseline (84.28 vs 78.85, so 5.43 points), meaning they started off with more severe difficulties on the primary outcome measure. This is worth noting because baseline differences between groups can sometimes partly contribute to the size of the improvement seen after treatment. The reported advantage of active stimulation over sham was a reduction of 6.25 points after treatment and 6.17 points at one-month follow-up. The authors acknowledged this limitation. The authors have addressed several potential confounders, including expectancy effects (which me and my group showing two years ago that can explain some of the TMS effects) and the statistical analyses are generally strong, but a longer follow-up is needed to determine whether the benefits are durable. But this is a very good and promising start. We cannot expect already at this stage to examine what will happen, for example, 6 months or 12 months later. 

It is also important not to overstate what this study shows. The work supports further investigation of this approach as a potential addition to existing support, and its strength is that it uses a biologically direct approach, which may offer an important complementary route. More mechanistic research is now needed to establish how the stimulation affects the brain, whether the motor cortex is the optimal target, and whether more personalised approaches could produce stronger and more reliable benefits.

This trial provides evidence of statistically significant – but, importantly, rather modest and decidedly short-term improvements in social communication, with small effect sizes and a follow-up limited to one month. Any interpretation of the effects here is further constrained by the authors’ reliance on SRS-2 and the potential for expectancy effects in their study. In addition, while adverse events were mild, their higher frequency in the intervention group suggests tolerability should be interpreted cautiously, and, from my own experience, the choice of stimulation target – the motor cortex - also raises questions about mechanistic specificity. These findings are best interpreted as preliminary evidence of modest short-term effects, rather than support for clinical implementation at this stage. Finally, I am concerned about the use of brain stimulation in such a young population as we are still uncertain about its effect on the developing brain.