What causes autism, according to the evidence
Not one cause, and never a single moment that “went wrong.” Autism emerges from a layered interplay of genetics, brain development, and environment — here’s what the research actually supports, and what it firmly rules out.
The genetic foundation
Genetics is the single most well-established contributor to autism, with heritability estimates ranging from 40% to 80%. That doesn’t mean autism is inherited the way eye colour is — it’s far more distributed than that.
Autism is polygenic: hundreds of identified genetic variants each contribute a small effect, rather than any single “autism gene” acting alone. Most of these variants don’t cause autism directly — they influence how the brain wires itself in regions handling sensory processing, language, motor coordination, emotional regulation, and social cognition.
Some variants are inherited from parents; others arise as spontaneous, de novo changes in the egg or sperm, appearing for the first time in the child rather than being passed down. Rare genetic syndromes — Fragile X syndrome and Rett syndrome among them — can share overlapping traits with autism, though together they account for only a small fraction of cases.
How brain development shapes the picture
Neuroimaging shows autistic brains follow distinct developmental trajectories — some regions with increased connectivity, others with delayed or accelerated growth. These aren’t random glitches; they reflect a genuinely different way of organising and interacting with the world.
Sensory processing
Hyper- or hypo-sensitivity to light, sound, or texture, rooted in how specific neural circuits develop.
Motor coordination
Delays in crawling, walking, or fine motor skill often trace back to differences in movement planning circuitry.
Cognitive functioning
Difficulty with attention-shifting or transitions can reflect altered synchronisation between neural networks.
Emotional regulation
Many autistic people find it harder to recognise, process, or manage emotional states.
Language & communication
Some children show strong early nonverbal communication but delayed speech; others speak fluently yet struggle with social language.
Neuroscientist Dr. Sumantra Chattarji has described the brain as far from a fixed structure — its capacity to adapt through experience, or plasticity, means the right kind of early support can actively help it build more of the connections that serve a child well. Autistic brain development is dynamic, not fixed at birth, and remains capable of growth and learning throughout life.
Environmental factors, across three stages
“Environment” here doesn’t mean pollution alone — it spans the entire biological backdrop shaping early brain development, from before birth through the first year of life. None of these factors cause autism on their own; they modestly shift risk only alongside underlying genetic susceptibility.
Prenatal factors
- Maternal immune activation / infection
- Alcohol, tobacco, or substance exposure
- Toxic exposures (lead, mercury)
- Chronic maternal conditions (autoimmune, thyroid)
- Gestational diabetes, hypertension, obesity
- Certain medications (e.g. valproic acid)
- Low folic acid in early pregnancy
- Advanced parental age
Birth-related factors
- Preterm birth, especially before 37 weeks
- Low birth weight (under 2.5kg)
- Oxygen deprivation during labour
- Neonatal jaundice or early infection
Early postnatal factors
- Air pollution or heavy-metal exposure
- Very short intervals between pregnancies
- Serious infections in infancy
These are modest statistical associations, not deterministic causes — most children exposed to any one of these factors do not go on to be diagnosed with autism.
Myths about causation, corrected
Outdated ideas about what causes autism don’t just spread misinformation — they stigmatise parents and derail attention from what actually helps.
-
✕“Something must have gone wrong during pregnancy”
Autism isn’t the result of one missed prenatal vitamin or a single stressful week. It develops through complex biological processes, many beginning before conception — well beyond a parent’s control.
-
✕“Vaccines cause autism”
This traces to a single 1998 paper that was later retracted after its data was found fraudulent, with the author stripped of his medical licence. More than 25 large-scale studies since have found no link. The timing coincidence — many autistic traits emerge around the same age as routine vaccination — is exactly that: coincidence, not causation.
-
✕“Screen time or modern lifestyle causes autism”
Screens don’t cause autism, though heavy use can sometimes mask early signs or become a comfort behaviour in children already on the spectrum. Framing it as a parenting failure only adds guilt without adding insight.
-
✕“Older parents always mean autism”
There’s a real but slight statistical association between parental age and autism risk, tied to a modestly higher rate of spontaneous genetic variation. It’s a correlation, not a cause — most children of older parents are neurotypical, and many autistic children are born to younger parents.
-
✕“Trauma or neglect causes autism”
This echoes the long-debunked “refrigerator mother” theory from decades ago. Trauma doesn’t cause autism, though it can coexist with it or, in some cases, mimic autistic traits like withdrawal or speech delay — a pattern that too often leads to misdiagnosis, particularly in girls and women.
What genetic testing can and can’t tell you
There’s no single definitive genetic test for autism — but a few tools offer meaningful, if partial, insight.
| Test | What it does |
|---|---|
| Carrier screening | A blood test for prospective parents checking for variants linked to conditions like Fragile X or Rett syndrome, which can overlap with autism traits. |
| Genetic counselling | Helps interpret family history or an existing diagnosis, sometimes via chromosomal microarray analysis or whole exome sequencing. |
| Prenatal genetic testing | Used in higher-risk pregnancies to examine amniotic or umbilical material for de novo mutations, chromosomal anomalies, or overlapping syndromes. |
None of these tests can confirm whether a child will be autistic. What they offer is a clearer picture of risk and context — useful for informed preparation, not a verdict.
Where this leaves families
The causes of autism are genuinely as varied as its presentation — inherited variants, spontaneous mutations, and a web of subtle prenatal and early-life influences, working together rather than any single culprit acting alone.
Some families find real reassurance in understanding the science. Others find the uncertainty harder to sit with. Both reactions are valid — and neither changes what actually helps most: informed, consistent support tends to shape a child’s future far more than any test result ever could.
Frequently asked questions
No. Extensive global research, including more than 25 large-scale studies, confirms vaccines — including the MMR vaccine — do not cause autism.
A complex combination of genetic, neurological, and environmental factors — never a single trigger, and never lifestyle or parenting choices.
No, though trauma can co-occur with autism or mimic some of its traits, which sometimes leads to misdiagnosis — particularly in girls and women.
It’s linked to a modestly higher risk, particularly alongside genetic predisposition, but it isn’t a direct standalone cause.
No evidence supports this. Diet can influence behaviour and gut health once a child is autistic, but it isn’t a causal factor in autism itself.
The loss of previously gained skills, typically between 15 and 30 months, with exact mechanisms still under research — possibly involving immune responses, neurological shifts, or genetic factors.
