What is joint attention in autism?
Q: What is joint attention in autism?
A: Joint attention is coordinating attention with another person around a third thing on purpose, such as a child pointing at a dog and then checking that you saw it too.
It splits into responding to joint attention, following someone else's point or gaze, and initiating it, using your own gesture or gaze to share.
Is it the same as eye contact?
No. Eye contact is looking at a person; joint attention is sharing a third thing with them.
Which is harder in autism?
Initiating, more often than responding.
Why does it come first?
Words attach to shared moments, which is why we target it before vocabulary.
What are examples of joint attention?
Responding looks like a child following your point to a balloon; initiating looks like a child pointing, showing, or giving an object purely to share it rather than to get something.
The give-away is purpose: a request point wants the object, while a sharing point wants your reaction.
Does handing me a toy count?
Yes, showing and giving are classic initiating behaviors.
What about looking back and forth?
Coordinated gaze between object and adult is one of the earliest forms.
Which do you teach first?
Usually responding, then initiating, inside play the child already likes.
At what age does joint attention begin?
The traditional milestone places it toward the end of the first year, strengthening through the second.
Newer longitudinal work complicates that: 44% of infants had produced joint attention bids by 6 months and 92% before 9 months, with no sudden jump between any two months.
So is 9 months still the marker?
For consistent, frequent joint attention, yes; the earliest bids simply start sooner.
What if it is late?
Late emergence is a screening conversation, not a diagnosis.
When do you assess it?
At intake for every young child we see, before language goals are written.
Do autistic kids have joint attention?
Yes, but typically less of it and in a different pattern, with initiating affected more than responding. It is a difference of degree, not an absence.
In a study following autistic children for about a year, gestural joint attention predicted later language better than IQ or starting language level did.
Does it improve?
Yes, and it responds to intervention better than most early social skills.
Is less joint attention permanent?
No. It is one of the more teachable early targets.
How is it measured?
By counting bids and responses in play, which is how we baseline it.
Is lack of joint attention always autism?
No. Reduced joint attention appears whenever a child cannot easily perceive, process, or coordinate a social bid, and several causes have nothing to do with autism.
Hearing loss blocks responses to spoken bids, vision differences block responses to pointing, and global developmental delay reduces joint attention roughly in proportion to overall delay.
What gets ruled out first?
Hearing, every time, before anything else is concluded.
Can language delay alone explain it?
It can look similar, which is why evaluation covers more than speech.
Who sorts it out?
Your pediatrician and a developmental evaluation, not a therapy provider.
Why do autistic people struggle with joint attention?
The mechanisms are genuinely debated, and two explanations lead without excluding each other.
One is reduced social motivation, where social stimuli are less intrinsically rewarding; the other is attention-shifting cost, where switching between people and objects takes more effort.
Is one proven?
Neither. Both are supported candidate theories rather than settled fact.
Does the reason change therapy?
Not much; the teaching targets look similar either way.
What do you assume?
Nothing. We start from what the child does and build from there.
Why is joint attention important for language development?
Because a word needs somewhere to land, and shared focus is what tells a child which thing the word belongs to.
In a randomized trial of 58 preschoolers, a short daily joint-attention intervention produced larger expressive-language gains a year later than the control condition, with the biggest gains in children who started with the least language.
Does it predict outcomes?
It is one of the strongest early non-IQ predictors of later language in autism.
Does teaching it transfer?
Yes, gains showed up in language a full year after treatment ended.
Where does that fit here?
It is the reason communication goals in our plans start upstream of words.
How do you teach joint attention?
Through naturalistic, play-based approaches delivered by parents and therapists, of which JASPER is the most studied.
A 2026 meta-analysis of 18 randomized trials covering 1,165 children found joint attention improved with an effect size of 0.53, play skills 0.33, and autism symptom measures 0.70.
Does age matter?
Younger children improved more than older children in that analysis.
Which part responds best?
Responding to joint attention showed larger effects than other types.
What does it look like here?
Floor play in your living room, with parents coached to run it between sessions.
What do screeners weigh most heavily?
Toddler autism screeners lean on social-relatedness and joint-attention items, particularly the sharing point rather than the requesting point.
Children with autism often develop pointing to get something while not developing pointing to share something, and it is that distinction the items target.
Is one item diagnostic?
No. Screeners flag risk; only a full evaluation diagnoses.
When is screening due?
At the 18-month and 24-month well visits, per the American Academy of Pediatrics.
What if it flags?
Pursue the evaluation and start support in parallel rather than waiting.
Sources
Mundy, Sigman & Kasari, 1990, JADD, Kasari et al., 2008, Journal of Consulting and Clinical Psychology, Song et al., 2026, Journal of Autism and Developmental Disorders, Salter et al., 2025, Infancy, Hanen Centre: paying attention to joint attention, NICHD: when do children show symptoms of autism, American Academy of Pediatrics screening guidance.
Effect sizes are reported as published, with trial counts and sample sizes stated. Where mechanisms are debated, such as why joint attention differs in autism, we present the competing theories rather than picking one.