What causes autism?

Autism results from mitochondrial impairment during brain development. It is triggered by an immunological event (e.g., fever).

Summary

• The current consensus is that the origins of autism are complex and multi-factorial. However, there is a simpler explanation that encompasses all other theories. That is, autism originates in mitochondrial impairment.

• Mitochondrial abnormalities, either prenatally, during pregnancy, or in the earliest years of life when brain development is most critical, encapsualte multiple risk factors for a later autism diagnosis by several times.

The phone call that changed my science

It was May 8th, 2020 when I was sitting in my car alone in the parking lot outside a barber shop that had just been allowed to reopen during the COVID lockdowns. I was staring into the rearview mirror, admiring my new haircut -- the first I'd had in months -- when my phone rang.

A producer of The HighWire show was on the phone. He had read a viral article, to which I contributed, criticizing the lockdowns. The article presented data from the CDC demonstrating that COVID-19 was in community spread as early as Nov 2019. It was censored on social media, and the producer wanted me to appear on The HighWire to be interviewed by Del Bigtree so more people could hear the message.

I told him, "I will talk about lockdowns. I will talk about the CDC data. I will talk about disaster response and human resilience, but I will not talk about vaccines."

He said, "That's OK. You can talk about whatever you want."

At that time, there were hardly any tenured University faculty willing to go on record in opposition to the lockdowns, and he later told me that my appearance might help other scientists come forward with their views, despite threats of censorship and retribution at their home institutions.

He was right, but in that moment, my concern was that appearing with Bigtree would somehow discredit me as some kind of anti-vaxx nutjob.

So I asked him, "What is it with you and vaccines, anyway? Why are you so committed to being anti-vaxx?"

He then told me a story that changed my understanding of science, my career as a researcher, my views on vaccines, and my life.

He said, "I have two medically-documented, vaccine-injured, autistic sons. Both of them regressed into autism immediately after their vaccine injuries.

"They were both normal, engaging, outgoing boys until they were injured by vaccines."

"There has never been a real study as to whether vaccines cause autism," he told me. "And, in a way, it doesn't matter... no scientist, no clinical or epidemiological trial, no peer-reviewed journal article can deny my experience as a parent."

So I did The HighWire show.

There's no record of my appearance, as far as I know, because Bigtree and The HighWire were scrubbed from the internet a few months after my appearance. Nevertheless, when I reflect on how naive I was about censorship and the extent of persecution that was in store for those scientists that were critical of official policies, I know I would cringe if I were to rewatch myself. Since then, my views on vaccines, on COVID, and on science have changed.

I no longer treasure the randomized, double-blind, placebo-controlled, clinical trial. As I wrote in my Uncommon Cold (Seager 2024) book:

Nothing matters more than your N=1 personal experience.
- Seager (2024).

What I could not have known then was that many of the faculty, like Jay Bhattacharya, MD/PhD, who were once vilified, censored, persecuted, and pilloried by government officials and the mainstream media would later be vindicated and appointed to lead the institutions that were once used against them. Moreover, one of the most important goals in their new official capacities is to investigate the very question that The HighWire prompted me to confront more than five years ago.

What causes autism?

Increasing rates of autism diagnosis

Rates of autism spectrum disorder (ASD) diagnosis have increased exponentially in the last decade. For example, children aged zero to four years old were almost five times as likely to be diagnosed with ASD in 2022 than 2011 (Grosevenor et al. 2024). However, expanded screening and changes in diagnostic criteria contribute the increase (Lundström et al. 2022). While the temptation in epidemiology is to seek correlations between the diagnostic timelines and hypothetical causes, such as changes in vaccine status, or 5G wireless exposures, or microplastics, or mold, or any other hypothetical underlying cause of autism, these correlations are specious distractions that cloud investigation into the root causes.

The most informative cases with regard to etiology are those that are most severe. Sometimes referred to as profound autism, these are cases that are non-verbal, exhibit stereotypical behaviors such as stimming (soothing self-stimulation) and self-injury, experience difficulties in sensory-processing, and sometimes suffer from seizures. In these cases, diagnostic signals are strongest, and rates of false positive diagnosis low. However, timeline data on this subpopulation is unreliable, partly because changes in the way autism is now coded by clinicians makes it difficult to make intertemporal comparisons (e.g., Russell et al. 2022), and partly because what was once coded as some other intellectual or developmental disorder is now being coded as autism (e.g., Cardinal et al. 2021).

That means the epidemiological data regarding autism is useless for uncovering its origins. Most of the work in autism etiology suffers from a serious shortcoming. It lacks a mechanistic understanding of modes of action. That is, epidemiological correlations and associations over time do nothing to improve our understanding of brain development, neurodevelopmental regression, or the onset of ASD. As a consequence, most clinicians and scientists will say things like "The etiology of autism is complex, multifactorial, and difficult to decode. More research is needed."

The problem is that more of the same type of research that has failed to clarify answers to the most basic questions will only add to the confusion. It's no wonder that the public grasps for oversimplified explanations like "Vaccines cause autism," and who could blame them?

What I didn't know about Type 1 diabetes

In 2001 I had just completed my doctoral dissertation when my son was diagnosed with Type 1 diabetes. It was December, and he had a fever for a few days, like lots of kids do, but he didn't get better.

He got worse.

Finally, my wife called his pediatrician, who was also our next-door neighbor. She described our son's symptoms and at first the doctor didn't seem too concerned.

"It's probably just the flu," he said. "There's been something going around, and I've been getting a lot of calls about sick kids,"

But then my wife said, "Plus, he's peeing the bed. A lot," and that gave the doctor pause.

"Meet me in the emergency room right away," he said. "Your son has diabetes and your life is going to change."

The hospital was across the street from our house. I carried my son into the emergency room. I don't really remember what happened next. Someone must have taken him from my arms and admitted him to the Pediatric Intensive Care Unit.

They put him on an insulin drip. I sat next to him all night, practicing my injection technique with an insulin syringe and an orange so that when he was discharged, I would know how to administer the insulin shots that would allow his body to process carbohydrates.

Compared to a diagnosis of autism with neurodevelopmental regression, Type 1 diabetes seems easy. As long as I could monitor my son's blood glucose levels and calibrate his insulin injections for his diet and exercise, his prognosis would be excellent. Nothing else would really be required.

But that knowledge did little to relieve my grief and shame. I couldn't believe I didn't know what was happening with him, and I couldn't understand what might have caused it.

It wasn't until years later, when I was sitting in a dentist's office thumbing through the pages of an old Reader's Digest, that I read about the new results of a longitudinal study out of Finland that revealed the cause of diabetes. By tracking Vitamin D levels in pregnant women and their newborn children, Finnish researchers discovered that a deficiency of Vitamin D during the first year of infancy, when the immune system is undergoing its most critical developmental stage, predicts risk of later onset of Type 1 diabetes (Hyppönen et al. 2001).

My son was born in late October, in northern New York. Although my wife and I paid close attention to her diet to ensure sufficient protein intake while she was pregnant, we hadn't given a thought to her Vitamin D status, or his. In retrospect, it makes sense to me that there's no way he would have gotten enough sunshine during the first few months of his life, given our latitude and the season in which he was born. It would be no surprise to have learned he was Vitamin D deficient, but back when he was an infant I didn't know to check.

I wrote about the relationship between Vitamin D status in the first year of life and later onset of autoimmune disorders in my article Reordering Autoimmune Disorders. What I discovered is that all autoimmune disorders, including Type 1 diabetes, rheumatoid arthritis, multiple sclerosis, and Parkinson's, originate in deficiencies of Vitamin D during that critical stage of immune system development that takes place after birth.

The reason this remains hidden from medical researchers is that the onset of the autoimmune disease is disconnected from its origins by several decades. The Finnish study that revealed the cause of Type 1 diabetes began in the 1960's and wasn't published until 2021. That is, Vitamin D levels at the time of disease onset might be normal, but the critical determining factor is the level of Vitamin D during development of the immune system. Without Vitamin D during this critical stage, the immune system will not follow a typical developmental trajectory. Instead, it will be predisposed for onset of autoimmune disorder when activated by an immunological event, as in the case of my son's fever. None of this could have been made clear without the patient, fastidious, longitudinal work done over nearly half a century by the Finnish researchers. Even if we started new studies on other autoimmune disorders, the research required might not be available for at least another ten years. It is only after the fact that the conclusion seems obvious.

Mitochondrial dysfunction in autism

While Type 1 diabetes is an immunological disorder, autism is neurological. That is, diabetes results from atypical development of the immune system, while autism results from atypical development of the brain. Coincidentally, the immune system and the brain are in their most rapid stages of development at the same time, namely late-stage pregnancy and the first several years of life. It is likely that Vitamin D plays a role neurological development (Botelho et al. 2024), but its role in immunological development is much stronger. The critical factor in neurological development is mitochondrial function.

Just like immunological development relies on Vitamin D, neurological development relies on mitochondrial function.

There is a close association between autism spectrum disorder (ASD) and mitochondrial impairment. For example, fifteen years ago, a study of young autistic children in California discovered that up to 80% had blood markers indicating some degree of mitochondrial injury and/or metabolic disorder (Giulivi et al. 2010, Frye & Rossignol 2011). Since then, disrupted mitochondrial function in ASD patients has been revealed in several clinical trials and corroborated in animal models (Khaliulin et al. 2025).

Mitochondrial impairment during pregnancy is one of the strongest predictive factors of autism risk later in life. Every condition associated with gestational metabolic disorder, including pre-eclampsia, gestational diabetes, and placental insufficiency, more than doubles the risk of ASD later in the child's life. Moreover, use of valproate (an anti-epileptic drug sometimes used to treat migraine headaches) during pregnancy more than quadruples the odds of autism (Hoirisch-Clapauch & Nardi 2019) by disrupting mitochondrial energetics (Salsaa et al. 2020). Even prenatal environmental exposures that impinge upon mitochondrial function are implicated in increased risk of ASD (Frye et al 2021).

At least since 1998, some scientists have hypothesized that autism is a mitochondrial disease (Lombard 1998). Such speculation makes sense, given the intense metabolic requirements of the brain and the fact that neurons are among the cells most densely packed with mitochondria. Any disruption in healthy mitochondrial function is sure to impact brain function, because mitochondria perform essential executive functions during brain development in addition to ATP production. Healthy mitochondria are essential to guide the growth, differentiation, and pruning of neurons and their connections within the developing brain. When mitochondrial function is impaired, the neural physiology becomes atypical and the brain is more vulnerable to autism.

Many competing theories on the causes of autism, including mold toxicity, non-native electromagnetic frequency exposure, or heavy metal poisoning can be collected under the mitochondrial theory once researchers recognize that each of these environmental stressors acts through the mitochondria. Certain metabolic profiles are so closely associated with ASD in young children that blood panels may be used as diagnostic aids (e.g., Barone et al. 2018). However, I think there are two reasons why the etiology of autism is still so poorly understood:

1. Autism can emerge without evidence of acute mitochondrial impairment at the time of onset, because the critical mitochondrial insufficiency may have been present earlier in life during phases of intense neurological developmental,

2. Mitochondria dysfunction only primes the brain for autism by steering it towards atypical developmental trajectories. Clinical presentation of stereotypical autistic behaviors might not be revealed without some immunological activation (e.g., fever, illness, or vaccine) that initiates autism. Thus, an association between the immunological event and onset of autism can mask the root cause.

Both of these points of confusion can be resolved by a new theory of the origins of autism that I will present here.

Immunoactivition initiates neurodevelopmental regression

One of the tragic hallmarks of new diagnoses of autism is the sudden onset of neurodevelopmental regression (NDR) -- i.e., loss of developmental milestones that had previously been obtained, such as speech -- following an immunological activation. For example, vaccines activate the immune system and can cause neuroinflammation that modifies neurological operations. Additionally, the aluminum and mercury in vaccines are toxic to mitochondria and may present increased neurological risks (de Oliveira 2022). However, vaccines are not the only immunological event that can initiate onset of autism. When mitochondrial abnormalities are present, a fever can be sufficient to initiate NDR, even in the absence of a recent vaccination (Shoffner et al 2010).

Environmental insult: Air pollution

In the absence of acute immunological activation, such as vaccination or infection, the onset of autism with NDR may be particularly enigmatic. In such cases, acute environmental insult to mitochondria may be the initiating event. For example, Frye et al. (2022) found a relationship between environmental exposure to air pollution measured as concentrations of PM2.5 (suspended particles measuring less than 2.5 microns in diameter) an onset of NDR. It is already well-established that PM2.5 exposures adversely impact mitochondria. One possibility is that multiple environmental insults increase accumulated mitochondrial stress, such that any additional acute exposure overwhelms mitochondria that are already chronically impaired, resulting in an inflammatory response that initiates the onset of NDR.

Environmental insult: Acetaminophen (Tylenol)

William Parker, PhD has compiled an extensive bibliography suggesting that the pain reliever acetaminophen (Tylenol) is the cause of autism (WPLab 2025). The genesis of this hypothesis was likely a 2006 internet survey of parents with children who experienced adverse reactions to the MMR vaccine. The results showed that children who experienced illness (e.g., fever) following the MMR shot were eight times more likely to be diagnosed with ASD than those who did not experience illness. Because it is common for parents to give either acetaminophen or ibuprofen to reduce a child's fever, the survey also showed an association between use of these over-the-counter medicine and ASD. However, that statistically significant only in the case of acetaminophen (Schultz et al. 2008). Although I'm not convinced of anything by the survey results, Parker suggests several other lines of evidence that support the acetaminophen hypothesis. What he overlooks is that mechanism by which acetaminophen is toxic at high doses is via mitochondrial injury, and that mitochondria that are already impaired in function prior to acetaminophen exposure are most prone to toxic dysfunction (Ramachandran et al. 2023). Therefore, the increased risks of neurodevelopment regression associated with acetaminophen that Parker is correct to be concerned may be the result of the additional load placed on mitochondria that are already compromised and an immune system that is already activated (e.g., by vaccine, fever, or both).

There are other causal theories of autism, including mold toxicity, exposure to microplastics, genetics, and heavy metal toxicity. The principal advantage of the mitochondrial impairment theory is that encompasses all the others. Every single credible theory of autism I’ve investigated proposes a mode of action that acts through the mitochondria. That is, all of these theories may be correct, insofar as an accumulation of mitochondrial insults from any of them during a critical period of neurological development could later result in autism. However, only the mitochondrial theory explains them all.

In a subsequent article, I will describe how to prevent and treat autism, and introduce you to a parent who has successfully recovered typical neurological capacities in her toy boys that suffered serious regression in childhood.