Dr. Pașca on Autism, Organoids, and Stem-Cell Cures

What most people get wrong about “curing autism”

Most debates about autism start from a false premise: that “autism” is one thing.

In this episode’s framing, autism is a behavior-defined umbrella, not a single biological disease with a single cause, a single pathway, and a single fix. That matters, because it changes what “cure” even means.

One side of the public conversation is often about identity and neurodiversity, people who have autistic traits and live independently, work, build relationships, and may not want to be “fixed.” The other side is about families living with profound disability, children who may be nonverbal, have epilepsy, have intellectual disability, struggle with sleep night after night, and require lifelong care.

This perspective does not treat those as the same clinical problem.

The key insight is practical: if you lump every presentation into one bucket, you end up with arguments that go nowhere and treatments that do not match the biology.

Important: In the discussion, the focus of “curing” is aimed at profound autism, the most severe cases where impairment is high and support needs are lifelong. This is a different target than helping a high-functioning adult navigate social stressors.

Autism as a spectrum and as many conditions, not one

Autism is described here as a spectrum, but also as a category that hides enormous diversity. Clinically, it is diagnosed by observing behavior over time, not by a blood test, a brain scan, or a single biomarker.

That point is easy to miss, and it is one reason autism discussions often feel like people are talking past each other.

Behavior-defined, biomarker-free

The discussion emphasizes that autism, like many psychiatric diagnoses, is defined by patterns of behavior. In medicine, that is unusual. For a heart attack, biomarkers and imaging can confirm the diagnosis quickly. For autism, there is no equivalent “troponin for autism.”

This creates two downstream problems.

First, there is a disconnect between the label and the biology. Two people can meet criteria for autism and share very few underlying mechanisms.

Second, without biomarkers, it is hard to track whether a treatment is altering a core mechanism or simply changing surface behaviors.

Profound autism as a distinct clinical urgency

A practical theme in the episode is that the “why cure?” question becomes much less abstract when you are talking about severe disability.

Profound autism, as discussed, often overlaps with:

For families in this situation, “treatment” is not primarily about making someone quirky into someone typical. It is about reducing suffering and increasing safety, communication, and independence where possible.

Did you know? The episode cites autism prevalence as close to 3% of the general population, a striking change from when it was considered rare.

Why prevalence rose, what we can and cannot say

The prevalence increase is real in the data, but the episode is careful about claiming a single cause.

The most grounded explanation offered is that multiple forces likely stack together.

Some of those forces are administrative and clinical, not biological.

Others might be biological, but harder to prove.

Explanations that do not require a new biological cause

The discussion points to several contributors that can raise prevalence without any true increase in underlying incidence:

These are not “gotchas.” They are the reality of behavior-based diagnoses.

Environmental contributors, plausible but complex

The episode also acknowledges that environmental exposures can contribute in some cases. A historical example mentioned is thalidomide, a drug once given in pregnancy that is known to cause major birth defects and has been associated with increased autism risk.

But the overall message is restraint: environmental associations are easy to propose and hard to prove as causal, especially across a broad spectrum.

This is where the episode’s “fever analogy” becomes useful. In the 19th century, “fever” was treated as a disease. Now we understand fever as a sign with many causes, viral infection, bacterial infection, cancer, autoimmune disease, each requiring different approaches.

Autism, in this framing, is more like “fever” than like “measles.”

What the research shows: Modern surveillance reports also document rising diagnosed prevalence and shifting diagnostic practices. For background, see CDC’s Autism and Developmental Disabilities Monitoring resources on prevalence estimates and diagnostic patterns (CDC ADDMTrusted Source).

Genes, de novo mutations, and what “20% get a diagnosis” means

A major emphasis in the discussion is the strong genetic component of autism, particularly in severe forms.

Twin studies were highlighted historically as a turning point, moving the field away from outdated psychoanalytic theories like the “refrigerator mother” hypothesis.

Today, genetics is central, but it is not simple.

One punchy reality from the clinic: about 20% of families may leave with a genetic diagnosis.

That number is both hopeful and frustrating.

What a genetic diagnosis can look like

The episode describes several pathways:

A key nuance is that humans accumulate many new mutations. The challenge is not finding variants, it is identifying which ones are truly causal for a given child’s symptoms.

Why a gene name often does not change treatment yet

Even when a gene is identified, families may not receive a targeted therapy today. The genetic information may help with:

But a gene label is not automatically a treatment.

That gap, gene to mechanism to therapy, is where Pașca’s lab tools are positioned.

Pro Tip: If a child is diagnosed with autism and has seizures, developmental regression, or significant intellectual disability, ask the clinician whether genetic testing and epilepsy evaluation are appropriate. The right workup depends on the individual situation.

What kinds of genes show up in autism research

One reason autism is hard to “solve” is that associated genes do not sit in one neat pathway.

The discussion groups genes into functional categories, each hinting at different mechanisms:

This variety supports the episode’s central claim: autism is many conditions.

It also supports a practical implication. If two children have different categories of genetic drivers, a single behavioral therapy or a single drug is unlikely to address both at a mechanistic level.

Central versus peripheral origins

A particularly interesting edge raised in the conversation is whether autism-related traits could originate partly outside the brain.

The episode references elegant work in animal models suggesting that altering autism-associated genes in the periphery can still influence brain development and behavior. The idea is intuitive: if sensory systems or body signals are altered early, the brain may wire differently in response to an overwhelming or distorted sensory world.

This does not prove a single “peripheral cause” of autism in humans.

It does sharpen the question of what to measure and where to intervene.

The clinic reality today, supportive care and unmet needs

There is no single treatment for autism.

That is not a defeatist statement. It is a reflection of heterogeneity and the absence of a universal biomarker.

In the episode’s framing, current care is often about supporting function and quality of life, while science works toward targeted interventions for specific biological subtypes.

Here is what “treatment” often means in real life, especially for profound autism.

A subtle but important point in the episode is that improving “non-core” issues may produce dramatic changes in day-to-day behavior.

If a child sleeps poorly for a week, almost anyone’s social functioning would deteriorate. So, addressing sleep is not a minor add-on, it can be foundational.

»MORE: Consider creating a one-page “care map” for clinic visits, current therapies, sleep pattern, seizure history, communication level, triggers, and safety concerns. It helps clinicians see the whole picture quickly.

Fever, microbiome, sleep, and other “edge case” observations

This section is where the episode’s tone is especially practical: many ideas are plausible, some are overhyped, and most are not universal.

Fever and temporary symptom changes

Parents have reported that when a child with autism develops a high fever, certain symptoms may briefly improve, including language in some anecdotal accounts.

The episode treats this as intriguing but not definitive.

Several hypotheses are mentioned:

But the conclusion is restrained: it is anecdotal, not present in everyone, and autism is not one disease so a single explanation is unlikely.

Microbiome, plausible benefits without clear causality

The microbiome was discussed as an area that has generated excitement, including animal work and some human trials. The episode’s stance is cautious: improving gut health might improve quality of life, but that does not establish that microbiome differences cause autism.

This is an important distinction for families.

If a child has constipation, pain, or restricted diet, addressing GI health can matter a lot. But it is different from promising that a microbiome intervention will reverse core neurodevelopmental differences.

For background on where the science stands, the NIH notes active research into microbiome and brain connections while emphasizing complexity and the need for rigorous trials (NIH Human Microbiome ProjectTrusted Source).

Sleep as a high-impact, often neglected lever

A striking clinical detail in the episode is the estimate that 70 to 80% of profoundly affected individuals may have severe sleep disturbance.

That is not just inconvenient.

Sleep loss can worsen irritability, attention, learning, and caregiver stress. It can also complicate epilepsy management.

If you are a caregiver, it is reasonable to ask clinicians about:

For general sleep health context, the American Academy of Sleep Medicine and Sleep Research Society recommend adults aim for at least 7 hours of sleep per night for health, with pediatric needs varying by age (AASM consensusTrusted Source).

The bottleneck in psychiatry, the developing brain is hard to access

The episode’s core scientific argument is about why progress in psychiatric and neurodevelopmental disorders has been slow.

It is not only because the brain is complex.

It is also because the developing human brain is hard to study directly.

A comparison is made to cancer, where access to blood and tumor samples enabled molecular understanding and targeted therapies. Leukemias that were once overwhelmingly fatal in children became far more treatable as biology became measurable and testable.

For autism, schizophrenia, and related conditions, the organ of interest is not easily biopsied, especially during the prenatal and early postnatal windows when key wiring decisions occur.

That inaccessibility shapes everything:

This is the problem Pașca’s lab is trying to “shortcut,” by building human-relevant tissue models from stem cells.

Organoids and assembloids, what they are and why Pașca built them

Organoids can sound like science fiction, “a brain in a dish.” The episode tries to make them feel more concrete.

A brain organoid is a 3D cluster of human cells derived from stem cells that self-organizes into features resembling aspects of early brain development.

An assembloid (spelled “assemblid” in parts of the transcript) goes a step further by connecting different organoid regions or cell types to model interactions between circuits.

This approach is framed as a way to observe human developmental processes that are otherwise inaccessible.

What they can do well

Organoids and assembloids can help researchers:

A key theme is not that these models are perfect.

It is that they are useful in a domain where direct access is limited.

What they cannot do

It is equally important to understand limits.

Organoids are not a full brain. They do not recreate the entire body, immune system, endocrine environment, sensory experience, or long-term learning in a living world.

They also raise ethical and technical questions, which researchers actively debate and regulate.

For readers who want a grounded overview of organoids in neuroscience, the National Academies have discussed scientific promise and ethical considerations in emerging brain models (National Academies reportTrusted Source).

The “timer, not a clock” idea

A distinctive line in the provided key arguments is the notion that cells may keep track of developmental progression more like a timer than a precise clock.

That is a subtle but important scientific stance.

If development is governed by a sequence of molecular state changes that unfold over time, then understanding that machinery could allow researchers to:

The episode is transparent that mechanisms are still being worked out, and that is part of the point.

From model to medicine, what “cure” could realistically mean

The episode’s title uses the word “curing,” but the practical pathway described is incremental: define subtypes, map gene to mechanism, test targeted interventions, then translate carefully.

In this framing, a realistic “cure” might look like one of the following, depending on the subtype:

The key is specificity.

It is not “one cure for autism.” It is many mechanism-based treatments for many biological conditions that currently share a behavioral label.

A step-by-step view of how this approach could work

Expert Q&A

Q: If autism is strongly genetic, does that mean the environment does not matter?

A: The episode’s framing suggests genetics is a major driver, especially in profound autism, but it does not rule out environmental influences. Instead, it argues that “autism” includes many conditions, so the balance of genetic and environmental contributors may differ across subtypes.

In practical terms, addressing sleep, seizures, nutrition, and sensory stressors can still meaningfully improve day-to-day functioning, even when the underlying vulnerability is genetic.

Sergiu Pașca, MD, PhD (as discussed on Huberman Lab)

What to watch out for in the real world

Because stem cells are mentioned, it is worth being concrete about a safety issue without drifting into diagnosis or prescribing.

Stem cell “treatments” marketed directly to consumers for autism are often not the same as the research described here. The episode’s work is about modeling and mechanism-driven therapy development, not unregulated clinics offering injections with vague claims.

If you are considering any stem cell-related intervention for a neurodevelopmental condition, it is reasonable to ask:

For general guidance on evaluating clinical research and trials, the U.S. National Library of Medicine provides a public registry of clinical trials (ClinicalTrials.govTrusted Source).

Expert Q&A

Q: Why does a genetic diagnosis help if there is no targeted treatment yet?

A: In the episode, a genetic diagnosis is framed as a way to reduce uncertainty, connect families to syndrome-specific resources, and guide research participation. It can also alert clinicians to associated medical risks, like epilepsy or cardiac issues in certain syndromes.

It is also a foundation for future treatments, because mechanism-based therapies typically start with clearly defined biological subgroups.

Sergiu Pașca, MD, PhD (as discussed on Huberman Lab)

Key Takeaways