The Puzzle at the Centre of Human Evolution

Autism spectrum disorder (ASD) usually gets framed clinically — a developmental condition that shapes how people communicate, behave, and experience the world. But behind those definitions sits a deeper mystery: why is autism so distinctly human?

You don’t see autism diagnoses across gorillas or monkeys. Yes, animals can show quirks that look a bit similar, but it’s not the same thing. Which makes the question all the more fascinating: what is it about our brains that tips the balance towards autism in us and not them?

That’s where the new Stanford research, published in Molecular Biology and Evolution, comes in. Instead of chasing a mythical “autism gene” (spoiler: there isn’t one), the team zoomed out. They studied how different neurons evolved across species, and whether those evolutionary tweaks intersected with autism risk.

The headline result? The wiring that allowed us language, reasoning, and memory may also have nudged us closer to autism.

Following the Cells, Not Just the Genes

Led by Alexander Starr and Hunter Fraser, the Stanford group analysed over a million brain cells from six species using single-nucleus RNA sequencing. That let them drill down into gene activity in humans, chimpanzees, gorillas, marmosets and others.

Normally, evolutionary rules of thumb apply. Abundant cell types — the ones that make up the backbone of the brain — hardly change. Rare or specialist cells are the ones that tend to evolve quickly. It’s common sense really: you don’t mess with the main circuitry unless you’re asking for trouble.

Except, in this case, trouble arrived.

Layer 2/3 intratelencephalic (L2/3 IT) neurons — the cortex’s most common excitatory neurons — refused to play by the book. Rather than staying stable, they changed rapidly in humans compared with apes. These neurons knit together distant parts of the brain, and they’re essential to thought, planning, and communication.

Even stranger, the evolutionary shift involved turning down the activity of autism-linked genes. Or, in plainer terms: natural selection seems to have chosen to “quiet” them.

Selection, Not Random Drift

When scientists spot rapid evolution, there are usually two explanations. Either natural selection is pushing those changes through, or the guardrails have loosened and mutations are slipping past unchecked.

To figure out which, the Stanford researchers checked variation in gene expression among humans today. If constraints had just relaxed, you’d expect messy variation. But what they found was the opposite: L2/3 IT neurons looked strikingly consistent person to person.

That points to positive selection. In other words, evolution nudged these neurons in a specific direction — reducing expression of autism genes — because it conferred an advantage.

Why Autism Genes in Particular?

This is the kicker. Why on earth would natural selection pick out autism-linked genes for special treatment?

The team ran comparisons across thousands of gene sets tied to human traits and conditions. And the strongest signal came from the SFARI database, which houses genes closely linked to autism. In humans, these genes were dialled down more than in chimpanzees or gorillas.

To double-check it wasn’t down to environment — diet, stress, age, all the usual suspects — they built hybrid human–chimp cortical organoids in the lab. These mini brain-like blobs carried both human and chimp DNA and grew in identical conditions.

And guess what? The human DNA still drove lower activity in autism genes. That nails it as a genetic feature, not a situational one.

The Payoff: What Did We Gain?

Which leaves the obvious question. What did we, as a species, actually get out of this?

The researchers don’t pretend to know for sure, but they propose two possibilities.

First: slowing down brain development may have given humans more time for language acquisition and learning. Autism genes often touch on synaptic growth. Dialling them down could have stretched the developmental window, trading slower maturation for greater flexibility.

Second: it may have been about stability. Bigger brains mean more potential for imbalance between excitatory and inhibitory signals. Autism is often discussed in those terms. By reducing activity of certain autism genes, evolution may have kept our neural circuits steady — but it came with the side effect of greater vulnerability.

Whichever way you slice it, the trade-off is clear: our mental edge may have carried autism risk in tow.

Patterns Beyond Autism

The study didn’t stop there. Similar down-regulation appeared for genes linked to schizophrenia, another condition considered uniquely human in its full form.

That lines up neatly with other research showing that Human Accelerated Regions (HARs) — fast-evolving stretches of DNA since we split from apes — are hotspots for autism and schizophrenia risk variants.

Put together, it suggests neurodiversity isn’t some random accident. It may be baked into the very process that made us human.

Numbers to Ground It

Of course, theories only matter if they connect with lived reality. And the numbers are stark.

The World Health Organization estimates autism affects about 1 in 127 people worldwide, based on 2021 data. Earlier studies pitched it closer to 1 in 100 children.

In the US, the CDC reported that 1 in 36 eight-year-olds were diagnosed in 2020, climbing to around 1 in 31 by 2022. Importantly, experts attribute this rise mainly to expanded definitions, earlier diagnosis, and greater awareness — not a sudden change in human biology.

Meanwhile, in lower-income countries, autism often goes under the radar thanks to weak health infrastructure and cultural barriers. But the underlying biology is presumed to be universal.

Cutting Through the Tylenol Noise

As if autism wasn’t complex enough, public discourse has been muddied further by the paracetamol debate. Claims that taking the drug (or Tylenol in the US) during pregnancy causes autism have spread like wildfire, with Donald Trump himself giving the rumours oxygen.

The science? The WHO is clear: there’s no conclusive evidence of causation. The FDA has updated labels to flag a possible association, but that’s a precaution — not a verdict. No medical authority recommends avoiding standard use in pregnancy.

To be fair, the subtlety of that position doesn’t make for snappy headlines. And so fear spreads, often drowning out serious research.

Reframing the Narrative

What the Stanford findings don’t say is just as important as what they do. They don’t imply autism is a mistake in our DNA. Rather, they suggest autism is part of the evolutionary price tag of being human.

The very neural tinkering that brought us Shakespeare, Einstein and the internet also increased our susceptibility to neurodevelopmental variation.

This isn’t to gloss over the real challenges autistic people face, nor the need for support. But it does change the framing: autism as difference, not defect.

It also stresses why inclusive policies and education matter. If autism is entangled with what makes us human, then supporting neurodiversity isn’t generosity. It’s recognition.

The Unanswered Questions

Plenty is still unknown. How exactly do these regulatory shifts affect development? Why do some individuals with similar genetic profiles develop autism while others don’t? And how do environment and genes interact in this complex picture?

Future studies will need to cast a wider net, across populations and developmental stages, to see how these evolutionary changes play out in practice.

The Bottom Line

Science rarely serves up neat answers, and evolution almost never gives us something for nothing. This new study adds weight to an idea gaining traction: that autism is not just a condition to be explained away, but part of the same evolutionary machinery that made us human.

Takeaway: Our intelligence, creativity and social depth may be inseparable from our vulnerability to autism. Evolution handed us extraordinary brains — but not without strings attached.