Prediction in Autism Spectrum Disorder: A Systematic Review of Empirical Evidence.
Jonathan Cannon, Amanda M O'Brien, Lindsay Bungert and 1 others
PMID 33570249WHAT IT FOUND
Prediction differences in autism were mixed: brain responses to repeated tones and faces often showed less dampening, and social predictions were often impaired, but many learned predictions were intact.
The review did not conclude that therapy helps.
Key findings
01The evidence did not support a general prediction impairment; differences varied by task and by what had to be learned.
02Reduced dampening of brain responses to repeated tones and faces was a recurring finding, but it was not uniform across stimulus types.
03All four studies of predictions based primarily on social expectations reported some impairment.
STILL TO COME
How it was doneWhat they found
Read the rest of this summary
You get three full summaries a month, free, and we do not ask for a card. Search, the TL;DRs and your library stay unlimited either way.
What it does not show
The review did not perform a meta-analysis, so it does not provide a pooled estimate of how large the differences are. The search relied on prediction and expectation terms, so studies of related processes that did not use those words may have been missed. The authors had a potential stake in the hypothesis: the last author wrote one of the original prediction-in-autism papers, and all authors are active researchers in this area. The 47 included studies used controlled laboratory tasks, not naturalistic daily activities, so it is unclear how these differences affect everyday function. EEG differences were not clearly connected to behavioral differences. The studies varied widely in methods and did not relate prediction differences to autism symptom severity, phenotypes, or genetic profiles.
Declared interests
The authors declared no conflicts of interest. The publication types indicate research support from N.I.H. extramural and non-U.S. government sources.
The easy way to misread this
Do not conclude that autism is a general prediction impairment or that prediction-based therapy is supported. The review found mixed, task-specific differences in laboratory studies and did not conclude that training helps.