Movement Synchrony in Autism Belongs to Both Partners

What the research found

Parents often ask why their child seems out of step with everyone else. The question has a literal version, and researchers have spent a decade measuring it. When two people interact, their movements tend to align in time, and that alignment can be recorded with cameras and motion sensors and coded by computer rather than judged by eye. The name for it is interpersonal motor synchrony.

Sixteen experimental studies of synchrony in autism, published between 2013 and 2021, were gathered and reviewed by Laura Carnevali, Irene Valori, Giorgia Mason, Gianmarco Altoe and Teresa Farroni, developmental psychologists at the University of Padova and the Technische Universitat Dresden. Thirteen of them, covering 512 pairs of people, were pooled into a single statistical estimate, published in early 2024 in the journal Frontiers in Psychiatry.

Across those pairs, synchrony was higher when both partners were non-autistic than when one of the two was autistic. The pooled difference reached a Hedges’ g of 0.85, a value that would ordinarily be described as substantial. The authors decline to lean on it. Their confidence interval runs from 0.35 to 1.35, and they write that “although the estimated effect size could be considered large, the wide width of the confidence interval suggests that also small to medium effect sizes should be considered consistent with the observed data.”

The uncertainty runs deeper. A prediction interval, which estimates the range a future study might land in, extended from minus 0.89 to 2.60. It crosses zero, which means the evidence does not exclude the possibility that in some settings the difference shrinks to nothing or runs the other way. The studies also disagreed with one another to an unusual degree, with roughly 90 percent of the variation among them attributable to genuine differences between studies rather than to chance. With participants ranging in age from 2 to 51 years and too few studies to test for an effect of age, the pooled figure covers toddlers and middle-aged adults on the same line.

In an adult study that assigned partners roles, autistic participants synchronised during fewer periods when following someone else’s lead than when leading or when no roles were set, and their periods of synchrony were shorter overall. General motor ability explained part of that difference and not all of it. General social skill predicted nothing.

What it means for you and your child

Carnevali and her colleagues offer an interpretation that places the difficulty between the two partners. They suggest that the non-autistic member of a pair may be struggling just as much, since “the TD member itself may have difficulty synchronizing with the ASD member because there are differences in sensorimotor and cognitive functioning that do not facilitate decoding the signals that the other is sending.”

That is an argument about what the numbers might mean, and the authors present it as one. The pooled studies compared mixed pairs against non-autistic pairs, a design that can register a gap without showing which partner produced it. The review also notes a theoretical bias in the existing literature toward deficit-based accounts of autism, which shapes both what has been measured and how the measurements have been read.

A parent who has watched an autistic child fall into easy company with another autistic child will want to know whether autistic pairs synchronise well together. The question is open. One study made that comparison and was excluded from the pooled analysis, and the authors write only that “deeper investigations would shed light on whether sharing similar experiences of the world support synchronization with one another.” Anyone who states the answer confidently is ahead of the evidence.

Synchrony was measured between two people and cannot be assigned to one of them on this evidence, which is solid enough to carry into a school meeting or a clinic appointment. General social skill also did not predict how well an autistic participant synchronised, so a child who never quite finds a shared rhythm with a teacher or a clinician has told that adult very little about their social understanding.

None of this settles what produces the difference, and the plausible values run low enough that in many ordinary encounters the difference may be slight. What the review does fix is the location of the measurement. When your child and another person fail to find a common rhythm, the recording was made between two people, and nothing in it locates the difficulty in your child.