What Causes Autism? The Real Science Behind a Misunderstood Question

For decades, families have searched for a clean, simple answer to one question: what causes autism? The honest answer from researchers is that there isn't a single cause, and there never was. Autism is mostly written into a person's biology from before birth, shaped by hundreds of small genetic differences that work together. That's the headline. The rest is detail — and the details matter, because confusion about causes has hurt families for generations.

This guide walks through what science currently says, what it rules out, and why the older "three main causes of autism" framing doesn't fit the evidence.

The Quick Answer: What Causes Autism?

Research across decades and millions of children points to a clear picture:

• Autism is overwhelmingly genetic. Twin and family studies estimate heritability at roughly 80%, with most studies landing between 64% and 91%.
• There is no single "autism gene." Autism is polygenic — meaning many common gene variants, each with a tiny effect, combine to shape autistic development. Rarer gene changes account for a smaller share of cases.
• A smaller portion of likelihood comes from specific, well-documented biological factors — notably advanced parental age, extreme prematurity, and certain maternal infections during pregnancy.
• Many things people commonly worry about — vaccines, parenting, screen time, and acetaminophen — are not supported by the evidence as causes of autism.

That's the full picture in four lines. The rest of this post unpacks each one.

Autism Genes: The Strongest Piece of the Puzzle

Genetics drive most of the variation in who is autistic and who isn't. Twin studies — the most powerful tool researchers have for separating biology from environment — consistently show high concordance among identical twins, who share nearly all their DNA. A meta-analysis published in the Journal of Child Psychology and Psychiatry placed heritability between 64% and 91%, with most estimates clustering near 80%. A reanalysis of Swedish twin data landed at about 83%.

But there is a key fact that often gets lost: there is no one "autism gene." Researchers have linked over 100 autism genes to autistic development, and the majority of cases are explained by the additive effect of many common variants — what scientists call a polygenic pattern. In a smaller subset (around 20% of cases), rare mutations like changes to CHD8, SCN2A, SHANK3, or copy-number variants at 16p11.2 can be identified. These same gene changes also show up in conditions like intellectual disability and epilepsy, which is why autism so often overlaps with them.

Is Autism From Parents?

This is one of the most common questions families ask. The short version: autism is from parents in the sense that the genetic variants shaping autistic neurology are usually inherited — but it is almost never traceable to a single parent or a single gene. Both parents contribute many small variants that combine in the child. In some cases, the variants are "de novo" — meaning they arose spontaneously in the egg, sperm, or early embryo and weren't carried by either parent at all.

So the question "is autism from parents" doesn't have a clean yes or no. The genetic material comes from parents, but autism isn't something a single parent "passes down" the way some single-gene conditions work. There is no parent to blame and no "carrier" to identify.

The Smaller, Non-Genetic Piece: What the Evidence Actually Supports

About 20% of the picture comes from non-genetic factors — and researchers can be specific about which ones. The factors with the most consistent, replicated evidence are:

Advanced Parental Age

A large multi-country study of nearly 5.8 million children found that autism likelihood rises with both maternal and paternal age. A separate meta-analysis of 27 studies found that each 10-year increase in maternal age corresponded with an 18% higher risk, and each 10-year increase in paternal age with a 21% higher risk. The effect is real but modest, and it's tied to biology — older eggs and sperm carry more spontaneous mutations — not to anything parents do or choose.

Extreme Prematurity

Children born significantly preterm — especially before 32 weeks — have an elevated rate of autism diagnosis compared with full-term births. Research from Children's Hospital of Philadelphia and large population studies place autism prevalence in the preterm population at roughly 7%, higher than the general rate. The mechanism is most likely tied to early brain development being interrupted by the medical realities of preterm birth.

Certain Maternal Infections During Pregnancy

Multiple meta-analyses, including a 2021 synthesis of 36 population studies, have found a small but consistent association between prenatal infection — particularly infections that cause fever or require hospitalization — and autism likelihood in offspring. Researchers believe the maternal immune response, not the infectious agent itself, is the relevant factor.

These factors raise probability somewhat. None of them "causes" autism alone, and none of them indicates anyone did anything wrong.

What Doesn't Cause Autism: Myths the Evidence Rules Out

Several theories about autism's causes have been studied extensively and not supported by research. Naming them matters, because misinformation about autism has caused real harm to families for decades.

Vaccines. This is the most-studied non-genetic factor in autism history. More than 40 high-quality studies across at least seven countries, involving over 5.6 million people, have found no link between vaccines and autism. A landmark Danish study reviewed the records of all 537,303 children born in Denmark between 1991 and 1998 and found no difference in autism rates between vaccinated and unvaccinated children. The 1998 paper that originally sparked vaccine fears was retracted from The Lancet, and its lead author lost his medical license.

Parenting. The "refrigerator mother" theory from the 1950s and 60s — which claimed cold parenting caused autism — has been fully discredited for more than fifty years. Twin and family studies starting in the 1970s made it clear that autism is a neurodevelopmental condition rooted in biology, not in how a child is raised. No parenting style causes autism. No parenting style cures it either.

Screen time. Early studies found small associations between toddler screen use and autism diagnoses, but those associations largely disappear when researchers control for socioeconomic factors and genetics. A 2025 analysis published in JAMA found that the screen-autism link vanished once family income and parental education were taken into account. Other research suggests the relationship may run the opposite direction — autistic children often seek out screens because they offer predictable, controllable sensory input.

Acetaminophen during pregnancy. A 2024 Swedish study of 2.48 million children — the largest of its kind — used sibling-controlled analysis (the gold standard for separating drug effects from family genetics) and found no increased autism risk (hazard ratio 0.98). Earlier studies that suggested a link likely reflected genetic factors shared within families, not a drug effect.

Why the "Three Causes" Framing Doesn't Fit

Autism isn't "caused" — in the everyday sense — by three things, or any specific number. It's the result of a developmental pattern shaped mostly by many autism genes acting together, with smaller contributions from a handful of biological risk factors. The question of whether autism is from parents misses the more accurate question: how do many small genetic variants combine across a family to shape who someone is?

For families, this matters because it shifts the conversation away from blame and toward something more useful: recognizing autistic traits early, accessing support, and building skills that help an autistic child or adult thrive in the world they're in. Autism is a way of being wired. A lifetime is plenty of time to learn, grow, and flourish in that wiring.

Supporting Autistic Children and Families Across Maryland and Virginia

At All Star ABA, we don't focus on causes — we focus on what comes next. Our Board Certified Behavior Analysts and Behavior Therapists work alongside autistic children and their families on the skills, communication, and independence that matter day to day. We serve families across Maryland, including Baltimore, Montgomery County, Annapolis, Columbia, Frederick, Carroll County, Pikesville, Owings Mills, and Washington County — plus parts of Virginia.

If you've been searching for honest answers and a clear next step, skip the late-night Googling. Reach out to our team for a no-pressure conversation about your child, your questions, and how in-home or community-based ABA might fit your family. We accept Medicaid and most insurance plans, our staff is bilingual, and there is no waitlist.

Frequently Asked Questions

Sources:

https://acamh.onlinelibrary.wiley.com/doi/10.1111/jcpp.12499

https://www.chop.edu/news/autism-and-prematurity-what-we-know

https://www.jci.org/articles/view/201157

https://www.nature.com/articles/mp201570

https://onlinelibrary.wiley.com/doi/10.1111/acps.12666

https://www.nih.gov/news-events/news-releases/study-reveals-no-causal-link-between-neurodevelopmental-disorders-acetaminophen-exposure-before-birth

https://www.cspi.org/article/vaccines-dont-cause-autism-heres-how-we-know