LYFsmile Logo
Early Signs of Autism in Babies: Can It Be Detected at 3 Months?
research-studiesApr 25, 2026|8 min read|Yakshi Shakya

Can a Baby Fitness Tracker Spot Autism Signs at 3 Months? New UCLA Study Aims to Find Out

A $3.1 million UCLA-led study could transform routine pediatric checkups, offering parents a powerful new window into early motor signs of autism spectrum disorder (ASD).

For most parents, the first concerns about autism arise when a child isn't talking by age two. But what if tiny, wearable movement sensors could flag subtle red flags as early as three months old – months or even years before speech delays become obvious? That's the promise of a landmark five-year study launched by UCLA Health researchers, funded by a $3.1 million grant from the National Institute of Neurologic Disorders and Stroke. The study, which started in January 2026 and runs through December 2030, is fitting 120 high-risk infants (baby siblings of children already diagnosed with autism) with comfortable, watch-like sensors on their wrists and ankles.

Why this matters for your family: Most autism diagnoses happen around age 4, after critical brain development windows have passed. This study aims to catch early motor signs – like coordination and grasping difficulties – that current routine checkups routinely miss because they only track major milestones (sitting, crawling).

Early detection means early intervention, when the brain is most plastic and responsive," says lead investigator Dr. Rujuta Wilson. "We're not waiting for language to emerge. We're reading movement patterns from day one.

Why Motor Signs Come First

Autism is often thought of as a social or communication condition. But research over the past decade shows that movement differences can appear before spoken words or eye contact delays. Here's the simple explanation:

  • Brain development is connected – the same neural circuits that control smooth, coordinated movement also help with attention, social engagement, and later language.

  • Babies learn through movement – reaching, grasping, and turning their head help them explore faces, objects, and sounds. If those movements are slightly irregular, the baby gets less practice with social feedback loops.

  • Motor delays at 3 to 6 months are measurable – long before a toddler fails to respond to their name or point at objects.

That's why this study focuses on movement variability – not whether a baby can roll over, but how they roll. Is it stiff? Jerky? Asymmetrical? Those tiny clues add up.

What the Sensors Actually Measure

The wearable devices look like soft fitness trackers or baby-safe sleep monitors. They are placed inside comfortable warmers worn on wrists and ankles. Parents go about their normal day – feeding, playing, tummy time – while the sensors collect data at home.

Here's what the machine learning algorithm analyzes, explained in plain English:

  • Acceleration patterns: How smoothly the baby reaches for a toy.

  • Limb coordination: Whether left and right arms move together or one lags behind.

  • Grasping force consistency: Does the baby grip a rattle the same way each time?

  • Repetitive motion bursts: Unusual arm flapping or stiffening during play.

  • Movement variability: Is there enough natural messiness in movement, or is it too rigid and predictable?

Earlier UCLA research found that combining these metrics could predict ASD with high accuracy – well before standard behavioral checklists could.

How the Wearable Sensors Work (No clinics, no stress)

Parents don't need to bring their baby to a lab. The sensors are placed inside soft, comfortable warmers – like tiny fitness trackers for babies – and capture natural home movements at 3, 6, 9, and 12 months of age. Researchers also conduct behavioral assessments at each visit, with formal ASD evaluations at 12 and 24 months.

Then, machine learning analyzes movement variability – how smoothly, consistently, and purposefully a baby reaches, grasps, or coordinates their limbs. Earlier UCLA research found that these motor metrics can predict ASD with high accuracy, long before social or communication symptoms typically appear.

Who This Helps Most (Real-Life Scenarios)

Parents with one child already on the autism spectrum: Younger siblings (high-risk infants) get early monitoring without repeated clinic visits.

Families in rural or underserved areas: At-home sensors remove travel barriers to specialty autism centers.

Pediatricians in busy practices: A 15-minute sensor-based screening could replace vague "wait and see" advice.

Early intervention therapists: They get objective movement data, not just parent observations or video clips.

Researchers studying brain development: Large-scale, real-world movement data improves machine learning models for all children.

Note: The study currently focuses on high-risk infants (siblings of children with ASD). Future phases may test general population screening.

From Research to Real-World Pediatric Rooms

The ultimate goal is not just better science – but a scalable, affordable tool for pediatricians. Imagine a 15-minute checkup where a baby wears soft sensors while playing on a blanket, and the algorithm flags "movements outside typical range" – prompting early referral to therapy.

That could mean:

  • Earlier support for motor delays (occupational or physical therapy starting at 6 months, not 2 years)

  • Better outcomes in independence, social skills, and coordination

  • Less stress for families who currently wait months or years for a diagnosis

Most assessments happen at home, making the study accessible to diverse families. And parents receive direct feedback on their child's motor development – a rare benefit in research studies.

Potential Limitations (What This Study Does Not Do Yet)

It's important to be clear about what this research does not claim:

  • Not a diagnosis tool yet – The sensors provide risk flags, not a yes-or-no diagnosis. ASD evaluation still requires clinical observation.

  • Not for all babies – Initial testing is only on 120 high-risk infants. Larger trials are needed before general use.

  • Does not replace pediatrician judgment – Movement data is an additional layer, not a replacement for developmental screening (like M-CHAT).

  • Not a cure or treatment – Early detection only helps if early intervention is available. The study does not provide therapy itself.

Researchers are aware of these limits and plan follow-up studies on cost, scalability, and accuracy across diverse populations (different races, income levels, and geographic areas).

How You Can Participate or Prepare Now

If you are a parent or caregiver:

  1. Check if you qualify – The UCLA study is recruiting families with a newborn (under 3 months) who has an older sibling with confirmed ASD. Contact UCLA Health's autism research division.

  2. Talk to your pediatrician – Even without sensors, you can ask: "Please track my baby's movement quality – not just major milestones – at 3, 6, and 9 months."

  3. Watch for early motor signs at home (see list below).

If you are a pediatrician or therapist:

  • Consider integrating movement variability checklists into well-child visits.

  • Stay updated on wearable validation studies – clinical adoption may begin in late 2027 if results are strong.

If you are a researcher or journalist:

  • Follow Dr. Rujuta Wilson (UCLA) and the National Institute of Neurologic Disorders and Stroke for publications in 2027 to 2028.

At-Home Motor Signs Parents Can Notice (Without Sensors)

While the study sensors are precise, parents can look for these general patterns starting at 3 to 6 months:

Typical movement: Reaches for dangling toys.
Possible early motor sign to note: Rarely reaches, or only uses one side.

Typical movement: Grasps and shakes a rattle.
Possible early motor sign to note: Stiff grip, doesn't adjust hand position.

Typical movement: Rolls from back to tummy.
Possible early motor sign to note: Jerky, asymmetrical, or avoids rolling.

Typical movement: Brings hands to midline (clapping position).
Possible early motor sign to note: Hands stay at sides or behind back.

Typical movement: Varies movement during play.
Possible early motor sign to note: Repeats the same arm flap or leg kick over and over.

Important note: These signs alone do not mean autism – but they are worth mentioning to your pediatrician, especially if there is a family history.

Key Takeaways (One-Minute Summary)

What's new: Wearable sensors (like baby fitness trackers) measure movement variability at 3 to 12 months.

Why it works: Motor delays often precede language delays in ASD.

Who it's for: 120 high-risk infants (siblings of children with autism) in a UCLA study funded by a 3.1 million dollar grant.

Biggest benefit: Early flagging leads to early therapy leads to better outcomes in independence and social skills.

Current limits: Not a diagnosis; small sample size; not yet available in pediatric offices.

What you can do now: Ask your pediatrician about motor quality; watch for asymmetrical or repetitive movements at home.

Final Word

This UCLA study is not science fiction – it is a 3.1 million dollar, five-year real-world trial happening now. If successful, the same technology could one day be as common as a baby scale or a thermometer in a pediatric exam room. For parents who have anxiously watched and waited for words that never came on time, this is a shift from wait-and-see to measure-and-help. Early help starts with early seeing. Movement speaks before words.

Need professional help?

Feeling suicidal or in crisis? Contact a helpline or emergency service immediately.

1. Vandrevala Foundation Helpline:
+91 9999666555 (24x7)

2. Sanjivini (Delhi-based):
011-40769002 (10 am - 5:30 pm)

3. Sneha Foundation (Chennai-based):
044-24640050 (8 am - 10 pm)

4. National Mental Health Helpline: 1800-599-0019

Newsletter

Get the latest mental health news delivered to your inbox.

Unsubscribe anytime. Privacy Policy

If you are in a crisis or any other person may be in danger - don't use this site.
These resources can provide you with immediate help.

LYFSMILE

With Lyfsmile, you can easily schedule online counselling sessions with the best psychologists, counsellors, and therapists in India. With over 120K clients in 70+ Countries, Lyfsmile is providing a safe, secure and confidential space to the clients.

DISCLAIMER

We are not medical healthcare provider or a hotline for suicide prevention. Call a suicide prevention hotline right away if you are experiencing suicidal thoughts, or go to the hospital.

Head Office

B710, Sushant Lok Phase I, Sector 43, Gurugram, Haryana 122007

Branch Office

Plot No 96-A , Block - B , Sector -13, Dwarka, New Delhi -110078

© 2019 - 2026 Lyfsmile | All rights reserved.