Most Trainings Helped, But the Gains Stayed Put

What the research found

Autistic children often struggle with what psychologists call executive function. This is the set of mental control skills we use to hold information in mind, resist a distraction, and switch flexibly between tasks. Difficulty here can make daily independence harder, and it tends to spill over into social life. This review asked a practical question: can training these skills directly, often through computer games, actually improve them?

The authors, led by Angela Pasqualotto and colleagues at the University of Trento, gathered every randomised controlled trial and quasi-experimental study they could find on the subject. They began with 2,601 records and narrowed them, through successive screening, to 19 studies that met their standards. Together these studies had followed 705 children and young people diagnosed with autism, aged from 4 to 20 years. As is typical in autism research, the sample was heavily male: the male-to-female ratio was roughly five to one, with boys making up more than 80 per cent of participants.

The interventions fell into two broad families. Some were computerised, delivered through serious games, attention-training software, or movement-based games. Others were not, using paper-and-pencil exercises, behavioural techniques such as modelling and reinforcement, or mindfulness-based approaches. The review looked at two kinds of outcome: whether training improved executive function itself, and whether it reached further, easing autism symptoms and improving social skills.

On the first question, the news was largely encouraging. In the authors’ own words, “most of the interventions ... were effective in enhancing EFs.” Children who trained tended to get better at the specific mental skills they practised, such as attention, working memory, and holding back an impulse. The review concludes there is “growing evidence for overall effectiveness of EF training, particularly when computerized.” That last qualifier matters: the computer-based programmes showed the more consistent results.

The second question is where the picture grew cautious. The authors report only “limited evidence” that gains generalised to untrained skills such as social ability, and equally limited evidence that improvements lasted. Most studies measured children only briefly after training ended, within a window of one week to one month. So the review can say that these programmes often move the needle on the trained skill, in the short term. It cannot yet say that they change a child’s wider life, or that the change endures.

The authors are candid about why caution is warranted. They point to uneven study quality, a shortage of replication, small samples, and training protocols that differed so much from one study to the next that findings were hard to pool. Notably, none of the 19 studies included autistic children with intellectual disability, so the results speak only to a particular slice of the autistic population.

What it means for you and your child

If your child finds it hard to concentrate, hold instructions in mind, or shift gears, the honest headline here is modestly hopeful. Structured cognitive training, especially the well-designed computerised kind, can improve those specific skills for many autistic children. That is a real finding, drawn from the best evidence available, and it is worth knowing.

But hold that finding at the right size. The clearest thing these programmes do is make children better at the thing they practise. A game that trains attention tends to improve attention on similar tasks. What the research has not yet shown is that this carries over into the parts of life you probably care about most: making friends more easily, coping with change at school, managing a morning routine. In research terms this is called generalisation, and here the evidence is thin. A gain on screen is not the same as a gain in the classroom or the playground.

The question of durability is just as important. Because most studies checked in only a few weeks after training stopped, we simply do not know whether the benefits hold months or years later. It is reasonable to hope they might, but hope is not evidence. If you invest in a training programme, treat its effects as something to be maintained and revisited, not banked once and forgotten.

Keep in mind, too, who was studied and who was not. The children in these studies were overwhelmingly boys, and none had a co-occurring intellectual disability. If your daughter is autistic, or if your child has more complex needs, the findings apply less directly. This is not a reason to dismiss them, but it is a reason not to assume they transfer neatly to your situation.

A word on the nature of the evidence itself. This is a systematic review, which means it carefully summarises many studies rather than running a new experiment. That is a strength, because patterns that recur across studies are more trustworthy than any single result. But a review can only be as good as the studies beneath it, and the authors are frank that many of those studies were small, varied, and rarely repeated. So this is a promising signpost, not a settled conclusion.

For a family, the sensible posture is interested but unhurried. If a cognitive training programme is affordable, enjoyable for your child, and does not crowd out other supports, it may be worth trying, with clear eyes about what it can offer: a good chance of sharpening a specific skill, and no guarantee that the benefit spreads or lasts. Watch for real-world change rather than scores on a game. And treat any programme that promises to reduce autism symptoms broadly, or to deliver lasting transformation, with scepticism. The current evidence does not support that claim, and the most trustworthy researchers in this field are the first to say so.