Six months or more into this, you have probably heard some version of "40 hours a week, or you are wasting the window". It is said with confidence and it makes parents feel they are failing. The evidence does not say it.
Where the number came from
The 40-hour figure traces back to early intensive behavioural research from decades ago and has been repeated ever since, often by people selling hours. It became a rule of thumb, then a rule, then a source of guilt. It was never a finding that held up across studies.
What the evidence says now
In 2024 the largest meta-analysis of early childhood autism intervention pooled 144 studies and 9,038 children aged 0 to 8. It measured intervention amount three ways: hours per day, weeks of duration, and total cumulative hours. It found no robust evidence that any of these was associated with better outcomes, whatever the approach or the outcome measured. The authors wrote that professionals recommending interventions "should be advised that there is little robust evidence supporting the provision of intensive intervention". strong
The Cochrane review of early intensive behavioural intervention (EIBI) found weak evidence that it may help some children, and said the evidence mostly comes from small studies that are not well designed. moderate
A 2025 German trial went the other way on purpose. It tested a low-intensity, individualised naturalistic developmental program at roughly one and a half hours a week against early intervention as usual, in 134 toddlers and preschoolers. After one year it did not beat usual care on core social communication, though it may have helped repetitive behaviour and executive function. moderate Read fairly, that trial says two things: very low intensity is not magic either, and "usual care" that already includes some good support is a high bar.
In Quebec, the Institut national d'excellence en santé et en services sociaux (INESSS) re-examined intensive behavioural intervention (ICI) for preschoolers in 2024 and recommended offering "a diversity of evidence-based approaches and interventions, including ICI" rather than prioritizing ICI alone, with each child's needs assessed and response measured. official
So: no dose-response curve, weak evidence for the intensive model, and the province's own evidence body stepping back from a single-model default.
What hours can and cannot do
More hours can mean more practice. That helps if the practice is the right practice. It can also mean more fatigue, less sleep, fewer meals together, less unstructured play, and a child who learns that adults mostly want performance. None of those show up in an outcome measure, but you will see them at 6 pm.
There is probably a floor. A child who gets a good strategy once a month will not learn it. The research cannot yet say where the floor is. What it can say is that above some modest level, adding hours has not been shown to add benefit.
Fit: what to look at instead
| Signs of fit | Signs of mismatch |
|---|---|
| Your child moves toward the session, or at least settles into it | Your child cries at the door, or shuts down during it |
| Goals are about communicating, playing, participating, coping | Goals are about sitting still, eye contact on command, "quiet hands" |
| You are coached and given things to do at home | You wait in the hallway and get a progress sheet |
| Strategies show up in daily life within weeks | Skills exist only in the room, with that adult |
| The schedule leaves sleep, meals and free play intact | Therapy is the family's calendar |
| The team measures and adjusts | The team defends the plan |
If a program fits, a modest number of hours is enough for it to work. If it does not fit, more hours make the mismatch worse.
Where the disagreement is
The case for intensity. Some clinicians and researchers argue that the meta-analyses lump together very different programs and children, that studies rarely test the highest intensities properly, and that for children with high support needs, more structured teaching time produces skills they would not otherwise get. They point out that "no evidence of a dose effect" is not the same as "evidence of no dose effect", and that the intensive model has decades of clinical experience behind it.
The case against. Others, including many autistic adults and a growing number of researchers, argue that the burden of proof sits with the people asking a 3-year-old to work a full-time schedule. They note that the strongest long-term result in the field came from a low-hour, parent-delivered approach, that most intervention studies rely on unblinded caregiver ratings, and that harms are almost never measured. A program that has not tracked harms cannot claim to be safe at high doses.
Both sides agree on one thing: the decision should be individual, measured, and revisited. Nobody should be told a number before anyone has met their child.
What this means in Quebec
If your child is offered ICI through the public system, it is usually a set-format program (on the island of Montreal, one CIUSSS describes it as 12 to 15 hours a week for about 10 to 12 months, starting the year the child turns 4, with weekly parent sessions). official Confirm the details for your CIUSSS; formats vary and change. Accepting it is a choice, not a test of commitment, and you can ask what else the DI-TSA-DP program offers. See ABA: a fair page for how to judge the program itself.
If you are paying privately, the hour count is also a budget question. Money spent on a package that does not fit is money not spent on a speech-language pathologist, an occupational therapist, or respite. Money and benefits in Quebec covers what can be claimed.
Honest limits
This page argues against a number, not against intervention. Some children do well in structured, higher-hour programs and their parents are right to keep going. The point is that the hour count is not the evidence, the child's response is.
This week
- Write down, in one line each, what your child's current program is for. If the goals are about compliance rather than communication or participation, raise it at the next meeting.
- Look at last week's calendar and count hours of therapy, hours of free play, and hours of sleep. Decide if the ratio is one you chose.
- Ask your team: "How will we know in three months whether this is working, and what would make you change the plan?"
Sources
- Determining associations between intervention amount and outcomes for young autistic children: a meta-analysis — Sandbank et al., JAMA Pediatrics, 2024
- Early intensive behavioral intervention (EIBI) for young children with autism spectrum disorders (Cochrane review) — Reichow, Hume, Barton and Boyd, Cochrane Database of Systematic Reviews, 2018
- Complex, low-intensity, individualised naturalistic developmental behavioural intervention in toddlers and pre-schoolers with autism spectrum disorder: the A-FFIP trial — Freitag et al., Journal of Child Psychology and Psychiatry, 2025
- Autism intervention meta-analysis of early childhood studies (Project AIM): updated systematic review and secondary analysis — Sandbank et al., BMJ, 2023
- Parent-mediated social communication therapy for young children with autism (PACT): long-term follow-up of a randomised controlled trial — Pickles et al., The Lancet, 2016
- Évaluation de l'Intervention comportementale intensive (ICI) pour les enfants autistes d'âge préscolaire au Québec — INESSS, 2024
- Programme d'intervention comportementale intensive (ICI) pour les enfants de 4 ans — CIUSSS du Centre-Sud-de-l'Île-de-Montréal, checked 2026