Is Autism Inherited from One Parent? What Genetics Actually Shows

Introduction

If you've come here searching "which parent carries the autism gene," here's the most important thing to know first: there is no single autism gene. Treating autism inheritance as if one parent can be identified as "the carrier", like for cystic fibrosis or Huntington's, fundamentally misrepresents the genetics, can drive unnecessary blame between partners, and often leaves families with worse information than they started with.

This piece walks through what the genetics actually shows: autism is highly heritable, but polygenic (involving many genes), with both parents typically contributing variants, plus genuine de novo factors that aren't "inherited from" either parent. This is more complicated than a single-gene answer would be, but it's also less burdened by blame framing.

The Short Answer: Both Parents, Usually

The honest answer to "which parent carries the autism gene" is:

Autism is highly heritable (genetic factors account for an estimated 64-91% of autism risk, based on twin studies, Tick et al. 2016 meta-analysis)
But autism is polygenic, hundreds of genes each contribute small amounts of risk, with the cumulative pattern producing autism susceptibility
Both parents typically contribute genetic variants to autism risk
De novo mutations (genuinely new mutations not inherited from either parent) occur in some cases
Specific genetic syndromes with stronger inheritance patterns (Fragile X, Rett syndrome, tuberous sclerosis, others) account for a portion of autism cases
No single "autism gene" exists, the polygenic architecture means there's no clean test for an inherited carrier status

So when you ask "which parent carries it," the most accurate answer is usually "both, in different ways, with substantial individual variation."

What Polygenic Inheritance Actually Means

This is worth understanding clearly because the polygenic nature of autism shapes everything else about its inheritance.

Some genetic conditions follow simple Mendelian inheritance, a single gene with a dominant or recessive pattern that predicts outcomes for offspring.

Huntington's disease (one dominant gene), sickle cell disease (one recessive gene), and cystic fibrosis (one recessive gene) have clean inheritance math because they're monogenic.

Autism doesn't work this way. Research has identified hundreds of genes associated with autism susceptibility, and most autistic individuals have many of these variants in combination.

Some specific patterns:

Common variants of small effect — many genes each slightly increase or decrease autism risk
Rare variants of larger effect — some specific genetic variants substantially increase risk but are uncommon
Copy number variations (CNVs) — duplications or deletions of larger sections of DNA that can affect autism risk
De novo mutations — new mutations that arise in egg or sperm cells, present in the child but not inherited from either parent

Polygenic architecture means that having an autistic child usually reflects the accumulation of many genetic factors, most of them inherited from both parents, often with no single one being "the cause."

"But the autistic side of the family is clearly..."

Many families notice that one side of the family appears to have more autistic relatives than the other.

There are a few possibilities for what's actually happening:

One side may genuinely contribute more variants in a specific family. Given the polygenic nature, individual families can have uneven contributions from each parental line.

The other side may have undiagnosed autistic relatives. This is increasingly common. Autism diagnostic criteria have expanded substantially since the 1990s, and many older relatives (parents, grandparents, aunts, uncles) who would have been recognized as autistic with modern criteria weren't diagnosed in their era. This is particularly true for women, people of color, and those whose autism didn't match the original child-focused, male-template diagnostic criteria.

The dominant lineage may simply have more recognized autistic relatives because of when they were born. Younger relatives growing up post-2013 (DSM-5) are more likely to receive diagnoses than older ones who grew up under earlier criteria.

The pattern of "all autistic relatives are on my partner's side" often resolves under closer inspection, frequently with someone on the other side recognizing themselves in the autism description their child or grandchild received.

De Novo Mutations

Some autism cases involve genetic mutations that are genuinely new. They occurred in the egg or sperm cell, are present in the child, but aren't inherited from either parent. These are called de novo mutations.

A few honest things about de novo contributions:

They account for a meaningful but minority portion of autism cases
They appear more frequently with advanced paternal age (older fathers' sperm cells have accumulated more mutations)
When a child has autism due primarily to de novo mutations, neither parent "carries" the autism, the mutation occurred during reproduction
These cases aren't strongly predictive of recurrence in subsequent siblings (since the mutation wasn't inherited)

De novo mutations are one reason the "which parent carries it" framing breaks down: sometimes neither does in any meaningful sense.

Specific Genetic Syndromes with Stronger Inheritance Patterns

A subset of autism cases, perhaps 10-20%, depending on how the criteria are applied, are associated with specific genetic syndromes that have more identifiable inheritance patterns. The most common include:

Fragile X syndrome — X-linked inheritance, the most common identified genetic cause of autism
Rett syndrome — typically X-linked, with mutations in the MECP2 gene
Tuberous sclerosis complex — autosomal dominant inheritance
Phelan-McDermid syndrome — deletion of part of chromosome 22
Angelman syndrome — caused by changes affecting chromosome 15
Smith-Magenis syndrome — chromosome 17 deletion
And many others, each accounting for a small proportion of autism cases

For these specific syndromes, inheritance patterns can sometimes be identified, and genetic counseling can provide more specific recurrence information. For most autistic individuals, however, no such single syndrome is identifiable, and the polygenic explanation applies.

When Genetic Testing Makes Sense

Genetic testing for autism has improved substantially in recent years. The American Academy of Pediatrics 2024 guidance recommends chromosomal microarray analysis (CMA) as first-tier genetic testing for children with autism, sometimes followed by next-generation sequencing (NGS) for additional clinical information.

What genetic testing CAN do:

Identify specific genetic syndromes (Fragile X, Rett, others) when present
Identify copy number variations of clinical significance
Sometimes inform medical care decisions, particularly when specific syndromes have specific medical implications
Inform recurrence risk for future pregnancies in specific syndrome cases

What genetic testing CANNOT do:

Diagnose autism itself (no genetic test "confirms" autism)
Predict autism reliably in unaffected siblings or future pregnancies for most families
Identify "the autism gene" the child has (because there isn't one for most cases)

For more on whether genetic testing makes sense for your family, see our piece on genetic testing for autism.

What's Established About Environmental Factors

The original version of this article ventured into vague territory about "toxins and environmental pollutants" as potential autism causes. That framing is loose and opens space for misinformation. The honest version is more specific.

Established (with substantial research support):

Valproate (Depakote, an anti-seizure medication) during pregnancy — Christensen 2013 (JAMA) and replicated studies show a real increased autism risk, particularly with first-trimester exposure. This is one of the few environmental factors with established evidence.
Advanced parental age — both maternal and paternal advanced age are associated with slightly higher autism risk, with paternal age linked partly through de novo mutations.

Not established or actively disproven:

Vaccines — extensively studied; no link to autism. The 1998 Wakefield study was fraudulent and retracted. Multiple large studies (Madsen 2002, Jain 2015, Hviid 2019) involving over a million children combined have found no association. For more, see our piece on autism regression.
Vague "toxins and pollutants" — no specific environmental toxin or pollutant has been established as a cause of autism. Research continues, but the framing that autism is environmentally caused is contradicted by the strong heritability evidence.

The honest framing is that autism is primarily genetic (heritability 64-91%), with a small subset of cases having identifiable environmental contributions (notably valproate exposure), and most environmental "causes" being either unsupported or actively disproven.

When Parents Discover Their Own Autism Through Their Child

This is increasingly common and worth naming directly. Many adults, particularly women, particularly those who grew up under earlier diagnostic criteria, discover their own autism only after their child receives a diagnosis.

The pattern often goes:

A child is evaluated for autism, often as a toddler or young child
Parent learns about autism in depth for the first time during evaluation and treatment
Parents recognize their own traits in the descriptions
Parent pursues evaluation as an adult, sometimes years later
Parent receives their own autism diagnosis

This isn't a failure of earlier diagnosis (which often wasn't accessible for that generation). It's the inheritance becoming visible through generational diagnostic patterns. For many late-diagnosed autistic adults, this realization is meaningful. It provides a framework for lifelong experiences and connects them to the autistic community.

If you're observing autism traits in your child and recognizing things in yourself, that's not a coincidence. The genetic component of autism does sometimes manifest exactly this way, across generations, becoming visible when one generation has access to diagnostic understanding, the previous generation didn't.

Sibling Recurrence

For families with one autistic child, considering whether to have another:

Recurrence in siblings is approximately 20% (per the most recent AAP Pediatrics 2024 guidance, Hyman et al.). This is significantly higher than the general population rate (about 1 in 31 children, per CDC ADDM 2025), reflecting the genetic component.

What this means practically:

About 80% of younger siblings of autistic children won't be autistic
About 20% will be
Sibling-of-autistic-child status warrants enhanced developmental monitoring (Texas ECI and similar programs are appropriate)
Recurrence risk doesn't increase substantially with additional autistic children in the family, it stays roughly in this range
For families with specific identified genetic syndromes, recurrence risk may be different and warrants specific genetic counseling

The Honest Framing for Families

The most useful framing on this topic isn't "which parent is responsible", it's "what's happening with our child and how do we support them." Some practical implications:

Don't blame either parent. The "which side did this come from" framing is destructive to family relationships and usually inaccurate anyway. Both parents typically contribute genetic variants. Sometimes, neither parent contributed in any meaningful sense (de novo mutations).

If you're an autistic adult recognizing this in yourself, that's information. It can connect you to the autistic community and frameworks that may have been missing from your own life. It doesn't mean you caused your child's autism. You contributed to their genetic makeup, like both parents always do.

For future reproductive decisions, talk to a genetic counselor. Particularly if specific genetic syndromes have been identified or if you have multiple autistic children, a genetic counselor can provide individualized information rather than generic statistics.

Focus on what supports your child. Genetic information is interesting and sometimes practically relevant, but it doesn't change what your child needs day-to-day. Speech-language pathology, occupational therapy, mental health support when relevant, school accommodations, and family understanding matter more than identifying which parent "carries" what.

Conclusion

Autism is highly heritable but polygenic. There's no single autism gene; both parents typically contribute; de novo mutations account for some cases; and the genetic architecture means there isn't a clean "carrier" answer to the original question implied. What does matter is supporting your child specifically, their unique strengths, challenges, and needs, rather than figuring out genetic blame.

At Steady Strides ABA, we work with autistic children across Texas and respect that families come from many backgrounds, with different genetic histories and family situations.

If you'd like to talk through what kind of support might fit your specific child, contact us for a conversation with a BCBA. For questions specifically about genetic testing, your pediatrician or a clinical geneticist is the most directly relevant resource.

Frequently Asked Questions

Is there a single autism gene that gets inherited?

No. Autism is polygenic. Hundreds of genes each contribute small amounts of risk, with the cumulative pattern producing autism susceptibility. There's no single "autism gene" that gets carried by one parent. Both parents typically contribute genetic variants, plus de novo mutations (genuinely new mutations) occur in some cases. The "which parent carries it" framing fundamentally misrepresents the genetics.

Can genetic testing tell us which side of the family autism came from?

Generally, no, particularly for most autism cases. Current genetic testing (chromosomal microarray analysis, next-generation sequencing) can identify specific genetic syndromes when present (Fragile X, Rett, tuberous sclerosis, etc.), accounting for perhaps 10-20% of autism cases. For the majority of autistic individuals, no specific genetic syndrome is identified, and the polygenic nature means there's no clean "this side contributed it" answer. Genetic counseling can provide individualized guidance for specific families.

What's the chance my next child will also be autistic?

About 20% per AAP Pediatrics 2024 guidance, significantly higher than the general population (about 1 in 31 per CDC ADDM 2025). This reflects the genetic component. About 80% of younger siblings won't be autistic, but the elevated risk warrants enhanced developmental monitoring. Early intervention services (Texas ECI for children under 3) can begin if signs emerge, even before a formal diagnosis. For families with specific identified genetic syndromes, recurrence risk may be different. A genetic counselor can provide individualized information.

Does it matter how old the parents were?

Slightly, yes, but the effect is modest. Both advanced maternal age and advanced paternal age are associated with slightly increased autism risk. Paternal age is linked partly through de novo mutations (older fathers' sperm cells have accumulated more mutations during life). The effect sizes are relatively small, and most older parents do not have autistic children. Parental age isn't a useful single predictor.

Can lifestyle changes during pregnancy prevent autism?

Largely no. The strong heritability of autism (64-91%) means most autism risk is established before pregnancy through genetic factors. The few established environmental risk factors, notably valproate exposure during pregnancy, are specific medical situations rather than lifestyle issues. Vaccines do NOT cause autism (this has been extensively studied, and the original 1998 paper claiming a link was retracted as fraudulent). Vague claims about "toxins" or "pollutants" causing autism aren't supported by current research.

I'm an autistic adult who just realized this through my child's diagnosis. What should I do?

This is increasingly common and meaningful. Several paths are useful: pursue your own autism evaluation as an adult (finding clinicians experienced specifically with adult autism evaluation matters, since presentations differ from childhood); connect with the autistic adult community (online and local); learn from autistic adult writers and resources (Devon Price, Sarah Hendrickx, ASAN, others); recognize that the framework gives you context for lifelong experiences. You didn't cause your child's autism — you contributed genetics, as every parent does. The recognition can be meaningful both for understanding yourself and for understanding your child.