Non Linear Learning - Rethinking Education for Neurodivergent Learners

196 Oxytocin, Autism, and how Autism Research works

Non Linear Learning - Rethinking Education for Neurodivergent Learners

Show Notes

In this episode, psychiatrist and researcher Dr. Eric Strobl joins Dr. Vaish Sarathy to talk about a new re-analysis of the SOARS-B trial on oxytocin and autism.

While earlier studies found no clear benefit, Dr. Strobl's fine-grained, item-level analysis using machine learning uncovered consistent evidence that oxytocin can enhance social-emotional reciprocity the ability to engage, connect, and respond in social contexts. Together, they discuss: Why most autism drug trials fail to show benefit

What "blunt outcome measures" really mean in clinical research

How machine learning can extract signal from noise in complex data

What oxytocin actually does (and doesn't do) in real life

How future studies could use more nuanced, individualized measures

Resources and Links Mentioned Study Discussed: Strobl E et al. (2024). "Item-Level Analysis Reveals Oxytocin Improves Social-Emotional Reciprocity in Autism Spectrum Disorder." Preprint

Original SOARS-B Trial: Parker KJ et al. (2017). "A Randomized Clinical Trial of Oxytocin in Children and Adolescents with Autism Spectrum Disorder." Journal of the American Medical Association (JAMA Psychiatry) Link

Related Reading:

Oxytocin and Social Behavior

On machine learning in psychiatry: Nature – Machine learning in mental health research

Key Takeaways Oxytocin may help but not for everyone. Its most reliable effect seems to be reducing social anxiety and improving comfort in social exchanges.

Measurement matters. "Blunt" outcome scales can bury meaningful results under noise. Item-level, data-driven analysis reveals nuance traditional methods miss.

Autism heterogeneity is real. The same outward behavior can stem from different root causes - so future trials need precision tools, not averages.

Hope through better science. New methods aren't about hype—they're about accuracy, compassion, and smarter research design.

About Dr. Eric Strobl Dr. Strobl is a psychiatrist and data scientist at the University of Pittsburgh who develops innovative machine-learning algorithms to uncover hidden structure in medical data. His current work focuses on autism, neurodevelopmental conditions, and the use of AI to improve clinical trial design.

About Dr. Vaish Sarathy Dr. Vaish Sarathy is a TEDx speaker, PhD chemist, educator, and mom to a non-speaking autistic teen poet with Down syndrome. She hosts the Non Linear Learning podcast and leads the Non Linear Education course for parents and educators who believe that every brain can learn, given the right way to teach.

Stay Connected Instagram: @drvaishsarathy

Free Guide: Turn ON Your Child's Learning Switch

Join the Non Linear Education Waitlist: Get on the list here →

More from Non Linear Learning - Rethinking Education for Neurodivergent Learners

Sep 30, 2026

Your child isn't choosing to melt down. By the time behavior shows up, the brain has already decided, and the thinking brain is out of the room. Dr. Lisa Riegel explains why schools most often miss this, and what to do instead.

Sep 7, 2026

Why spelled communication is not a guessing game:

Jul 21, 2026

We don't make a neurotypical kid prove they can read before we hand them a book. We just teach them to read.

Jul 1, 2026

You've talked to 10 therapists. Tried 20 strategies. And your child still goes from calm to crisis in 30 seconds - for reasons no-one understands. Understanding a body requires that we understand the sensory information coming in. And Dr. Dana Johnson - OT, PhD, and founder of the Spellers Center - tells us that there are not 5 but 8 senses. Dana joins Non Linear Learning for the 3rd time to talk about interoception: the eighth sense, the one almost nobody gets taught, and the one that might explain what's been happening in front of you all along. Listen along to learn: Why your child's meltdown appears so sudden

Jun 15, 2026

We talk a lot about what our students can't do. But something that educators have found over the years is that there is a scatter to the skills we see in our students. The wildly uneven profile that we see could actually be a blessing. Mark Talaga, counselor, speaker, and founder of the Center for Identity Potential, joins the podcast to talk about asynchronous development in gifted and twice-exceptional learners. And here's what hit me: we're working the same problem from opposite ends. Where there is a sharp strength, there is often a significant deficit, but the opposite is true as well! This conversation got under my skin in the best way. Tune in to this episode to learn: What "asynchrony" actually means in gifted and disabled learners

See all episodes →