The Deficit Is Broad, Not Sharp

What the research found

Executive function is the set of mental controls that let a person hold a goal in mind, plan the steps, switch approach when something changes, resist the obvious wrong move, and keep track of information while using it. Researchers have long asked whether people with autism have a particular weakness here, and if so, whether it is one sharp deficit or many. This 2018 paper in Molecular Psychiatry set out to answer that by pooling the existing evidence.

The authors combined 235 studies comparing people with autism spectrum disorder to neurotypical controls. That is a large body of work: 14,081 participants in total, 6,816 with autism and 7,265 controls. All participants were aged six or older; studies of younger children were left out. Pooling results this way is what a meta-analysis does. It does not run a new experiment. It takes many separate studies, weights them by size and reliability, and estimates the average effect across all of them.

The headline number is a moderate overall deficit in executive function. Using Hedges’ g, a standard measure of effect size, the pooled estimate was g=0.48, with a 95% confidence interval of 0.43 to 0.53. A g of about 0.48 means the average person with autism scored roughly half a standard deviation below the average control. That is a real, consistent gap, but it is a difference between group averages, not a wall between two populations. The two groups overlap heavily. Many individuals with autism score above the control average, and many controls score below the autism average.

The more striking finding is how even the deficit was. The researchers examined six domains: concept formation and set-shifting, mental flexibility, fluency, planning, response inhibition, and working memory. Effect sizes were small to moderate across all six, and the differences between domains were not statistically significant. In plainer terms, no single domain stood out as the problem. The weakness looked general rather than fractionated, spread thinly across the whole system rather than concentrated in one part.

The deficit also held steady across age. When the authors split studies by age group, the pattern was broadly stable from childhood into adulthood. The one exception was working memory, where younger participants showed no significant difference from controls. This does not mean executive function improves or resolves with age; it means the average gap, relative to same-age peers, stays roughly constant across development.

Two cautions belong with these numbers. The authors found evidence of publication bias, meaning studies with clearer results are more likely to be published, which can inflate a pooled estimate. A correction analysis, however, left the effect size essentially unchanged. Separately, and importantly for anyone hoping for a clinical tool, they reported that only a very limited number of executive-function measures were sensitive enough to distinguish individuals reliably. The average difference is real, but no single test cleanly identifies autism.

What it means for you and your child

If your child has an autism diagnosis, this research offers a fair expectation rather than a verdict. On average, children and adults with autism find executive tasks somewhat harder than peers do: holding instructions in mind, planning multi-step work, shifting when a method stops working, resisting the first impulse. The gap is moderate and broad. It is likely to touch several areas of school and daily life rather than one narrow skill you can name and isolate.

Because the deficit is general, support should be too. There is no evidence here that one domain, say planning or inhibition, is the master problem to fix first. A better reading is that scaffolding executive demands across the board tends to help: breaking tasks into visible steps, reducing what has to be held in memory at once, signalling transitions before they happen, and building routines so that fewer decisions have to be made under pressure. These are ordinary, practical adjustments, and the breadth of the finding is what justifies applying them broadly.

Hold the group average loosely when you think about your own child. A pooled effect of g=0.48 describes populations, not individuals. Your child may be well above it in some areas and below it in others. The heavy overlap between groups is the point: knowing the average tells you what is common, not what is true of any one person. Use your child’s own profile, from a proper assessment, as the guide to where support is actually needed.

Be honest, too, about what this study cannot say. It compares groups; it does not explain cause. It cannot tell you whether executive difficulty is a root feature of autism or a downstream effect of something else. The authors themselves note that they excluded reaction-time measures and questionnaire data, did not fully account for anxiety, and that anxiety in particular is strongly linked to poor executive performance. If your child is anxious, some of what looks like an executive weakness may ease as the anxiety is addressed. That is worth exploring rather than assuming a fixed limit.

Finally, treat any single test result with care. The researchers were clear that few executive-function measures are sensitive enough to stand alone, and that the field needs better, more ecologically valid tools. A one-off score is a snapshot under artificial conditions, not a measure of your child’s capability across a real week. The practical takeaway is steady rather than dramatic: expect broad, moderate executive challenges, support them broadly, watch for anxiety, and judge progress by how your child manages actual tasks over time, not by a number on a page.