How Does a Child Get Autism: Genetics and Risk Factors

Autism is not caused by any single event or exposure. It develops through a combination of genetic predisposition and prenatal environmental factors, with genetics accounting for the largest share of risk. Current estimates place heritability at around 85%, meaning the vast majority of what determines whether a child develops autism is written into their DNA before they are born. About 1 in 31 children in the United States is now identified with autism spectrum disorder, based on 2022 surveillance data from the CDC.

Genetics Play the Largest Role

Autism runs in families, and the genetic architecture behind it is complex. Rather than a single “autism gene,” the condition involves a mix of common genetic variations that individually have small effects and rare mutations that carry larger ones. These common and rare variations act together, adding up to create risk. This is why autism can appear in families with no prior history: a child may inherit a particular combination of ordinary gene variants from both parents that, together, cross a threshold.

The clearest evidence for genetic influence comes from twin studies. Identical twins, who share virtually all their DNA, show far higher rates of shared autism diagnoses than fraternal twins. When one identical twin has autism, the other is diagnosed the majority of the time. Fraternal twins, who share about half their genes, match at much lower rates, closer to what you’d expect for any pair of siblings.

Sibling data reinforces this picture. Research from the Baby Siblings Research Consortium found that about 20% of younger siblings of a child with autism will also be diagnosed. That rate climbs to nearly 37% in families where more than one older child already has autism. Boys face higher recurrence risk than girls: roughly 25% of younger brothers receive a diagnosis, compared to about 13% of younger sisters.

What Happens in the Developing Brain

During typical brain development, a baby’s brain produces far more connections between nerve cells than it will ultimately need. A natural process called synaptic pruning then trims away the excess, strengthening the most-used pathways and eliminating weaker ones. In children who develop autism, this pruning process appears to be insufficient across multiple stages of development. The result is a brain with more connections but less efficient communication between regions, which can affect how a child processes social cues, sensory input, and language.

These differences in brain wiring begin before birth and continue through early childhood. They are not something a parent causes or could have prevented through different choices after the child is born. The underlying blueprint is set by genetics and shaped by the prenatal environment during critical windows of fetal brain growth.

Prenatal Factors That Influence Risk

While genetics set the stage, certain conditions during pregnancy can nudge risk higher. None of these factors guarantee autism, and most children exposed to them do not develop it. But they are worth understanding.

  • Parental age. Both older mothers and older fathers are associated with increased autism risk. A large Danish study tracking over 1.3 million births found that when either parent was 35 or older, the risk rose measurably, with the highest risk when both parents were over 40. For mothers under 35, the father’s age still mattered independently, and vice versa.
  • Maternal infections and inflammation. Infections during pregnancy, particularly rubella and cytomegalovirus, as well as prolonged fevers, are linked to higher autism rates. The connection likely involves inflammatory signals that can cross the placenta and affect fetal brain development.
  • Certain medications. Valproic acid, an anti-seizure medication, carries the strongest known medication-related risk. It can interfere with gene expression during a critical window of embryonic brain development, disrupting how nerve cells form connections, migrate to the right locations, and communicate with each other.
  • Diabetes during pregnancy. Both pre-existing and gestational diabetes are associated with modestly increased risk.
  • Premature birth. Babies born preterm face significantly higher odds of an autism diagnosis. Extremely premature infants (born at 22 to 27 weeks) have an autism prevalence of about 6.1%, compared to 1.4% for babies born at full term. Overall, preterm birth raises the odds of autism roughly threefold.
  • Environmental exposures. Heavy exposure to air pollution (fine particulate matter), pesticides, heavy metals, and significant alcohol use during pregnancy have all been associated with increased risk. Folic acid deficiency during pregnancy is another identified factor.

How Environmental Exposures Affect Genes

One of the more important discoveries in autism research is that environmental exposures don’t just act alongside genes. They can change how genes behave. This field, called epigenetics, studies chemical modifications that turn genes on or off without altering the DNA sequence itself. Certain toxins, medications, and nutritional deficiencies during pregnancy can alter these chemical tags on a developing baby’s DNA, shifting which genes are active during brain formation.

Valproic acid, for instance, causes a specific type of change that leaves genes in an abnormally active state during embryonic brain development. Exposure to compounds like bisphenol A or lead alters a different set of chemical markers on DNA. Perhaps most striking, some of these epigenetic changes may be passed to future generations through modifications in reproductive cells, potentially affecting children who were never directly exposed themselves. The gut microbiome also interacts with this system: maternal diet, infections, and chemical exposures reshape the microbiome, which in turn can influence the epigenetic landscape of the developing fetus.

Vaccines Do Not Cause Autism

This question comes up often enough that it deserves a direct answer. Vaccines do not cause autism. A 2025 WHO analysis reviewed 31 primary research studies published between 2010 and 2025, covering data from multiple countries, and confirmed no causal link between childhood vaccines and autism. The same conclusion applies to the trace amounts of aluminum used as an adjuvant in some vaccines and to thiomersal, a preservative once common in multi-dose vials. The WHO has reaffirmed this finding repeatedly since 2002. A large cohort study using nationwide Danish registry data on children born between 1997 and 2018 reached the same conclusion.

The original 1998 study claiming a link was retracted, and its lead author lost his medical license for ethical violations and data manipulation. Decades of research involving millions of children across multiple continents have consistently found no connection.

When Signs Typically Appear

Although the biological roots of autism are present before birth, behavioral signs emerge gradually as a child reaches social and communication milestones. Some early markers are visible before a child’s first birthday: not responding to their name by 9 months, or not showing a range of facial expressions like happiness, sadness, or surprise by that same age. By 12 months, a child who doesn’t use gestures like waving goodbye or play simple interactive games like pat-a-cake may warrant closer attention.

Between 15 and 24 months, the signs often become clearer. Not pointing to share something interesting by 18 months, or not noticing when others are hurt or upset by age 2, are commonly noted markers. By age 3, not joining other children in play is a significant indicator. Later milestones include not engaging in pretend play by age 4 or not performing for others (singing, dancing, acting) by age 5. These milestones aren’t rigid cutoffs, but missing several of them is a reason to seek developmental evaluation.

Most children receive an autism diagnosis between ages 2 and 4, though the condition can be reliably identified as early as 18 months in some cases. Earlier identification generally leads to earlier support, which tends to improve outcomes for communication and social skills during the years when the brain is most adaptable.