Three Groups, Three Trajectories

What the research found

Executive function is the set of mental tools we use to hold a goal in mind, resist a distraction, and switch approach when the first one fails. It develops slowly through childhood, and it underpins much of what we call “learning to manage oneself” in a classroom. This study asked a narrow question: how does executive function unfold, over years, in children who are more likely than average to be autistic?

The researchers followed three groups of children. Eighty-three were at high likelihood of autism (they had an older autistic sibling) and were later diagnosed as autistic. Two hundred and thirty-five were at high likelihood but were not diagnosed. One hundred and thirty-two were at low likelihood, with no family history. The children were assessed at 12 months, at 24 months, and once more at school age, between five and fourteen years old. Not every child returned for every visit, which is normal in a study that runs this long.

Two kinds of measure were used. In infancy the researchers used a hands-on task in which a child watches an object hidden, then must resist reaching to where it used to be. At school age they used timed laboratory tasks of attention and mental flexibility, and separately asked parents to rate their child’s everyday executive function on a standard questionnaire, the BRIEF-2.

The results did not describe a simple decline. On the laboratory tasks, the two high-likelihood groups looked broadly stable across time. It was the low-likelihood group that stood out early: between 12 and 24 months their standardised scores rose (p less than 0.05), while neither high-likelihood group changed. In other words, the typical pattern was improvement, and the high-likelihood children, autistic or not, simply did not show that early gain. By school age the autistic children scored lower than both other groups on every measure (p at or below 0.050).

The parent-reported picture was different again, and it is the finding that matches the study’s headline. On parent report, the autistic children’s standardised scores decreased over time, the low-likelihood children’s increased, and the non-autistic high-likelihood children stayed roughly level. The authors describe this as “accumulating EF difficulties” in the autistic group. So there are two distinct signals here: a very early divergence on direct tasks, and a widening gap in everyday functioning as parents observed it across the years.

The authors are candid about the limits. They did not correct for multiple comparisons, which they note may have inflated the chance of a false positive, so the individual p-values deserve caution. The school-age band is wide, five to fourteen, which can blur age-specific patterns. The preschool years were not measured at all. And the analysis is at the group level, so it says nothing about any one child’s path.

What it means for your child

The useful takeaway is not “autistic children’s executive function declines.” That framing is too strong for what the data show, and it is the opposite of encouraging. On the direct laboratory tasks the autistic children were fairly stable; the more accurate description is that they did not show the early surge of improvement that low-likelihood peers did, and that everyday demands, as parents rated them, grew harder to meet over time. A gap can widen because one group falls or because the others pull ahead. Here the second reading fits at least as well as the first.

That distinction matters for what you do. The study’s own recommendation is modest and sensible: for children at high likelihood who are already showing early executive-function difficulties, supports beginning in toddlerhood may be worthwhile. It does not claim that early support prevents autism, changes a diagnosis, or closes the gap. This is an observational study. It can show that trajectories differ; it cannot show that any intervention would have altered them. No treatment was tested.

It is also worth holding the two measures apart in your own thinking. Timed laboratory tasks and a parent questionnaire are not the same thing, and here they told partly different stories. If your child performs adequately on a structured test but struggles with transitions, planning, or self-regulation in daily life, that combination is entirely consistent with what this study found. The everyday difficulty is real information, not a measurement error.

For a parent of a younger sibling of an autistic child, the practical stance is watchfulness without alarm. Early differences in executive function were visible in this sample by 24 months, before school and well before most formal assessments. That is an argument for paying attention to how a toddler holds a goal, shifts attention, and copes with change, and for seeking support early if those skills seem effortful, rather than waiting for school to surface the problem.

Finally, keep the scale of the claim in view. This is one well-run study of siblings, a group chosen precisely because they are not representative of all children. Its numbers are real and worth taking seriously. They are also a description of averages across groups, not a forecast for any individual. Your child is not a trajectory on a chart. The value of work like this is that it tells you where to look early, not what will happen next.