Autism spectrum disorder (ASD) has no single cause. It arises from a combination of genetic, prenatal, and early developmental factors that shape how the brain forms and functions. The latest CDC data from 2022 puts the prevalence at about 1 in 31 children aged 8, up from 1 in 150 in 2000. That increase reflects broader diagnostic criteria and better screening, but it also drives urgent questions from parents about what’s actually behind the condition.
Genetics Play the Largest Role
The strongest evidence points to genetics as the primary driver. Twin studies estimate that genetic factors account for roughly 80 to 87 percent of the variation in autism risk. When one identical twin has autism, the other is far more likely to have it than a fraternal twin would be, and the condition clusters strongly in families. A large study published in JAMA estimated broad-sense heritability at around 83 percent, with shared environmental factors contributing only about 4 percent.
There is no single “autism gene.” Hundreds of genes contribute, each with a small effect. Some involve rare mutations with larger impact. For example, having extra copies of a gene called UBE3A on chromosome 15 increases risk. Mutations in MECP2, a gene involved in regulating how other genes are read, contribute to related neurological conditions and also play a role in autism. The genetic picture is complex: some variants are inherited from parents, while others arise spontaneously in the child.
Brain Development Differences Start Early
In children with autism, something goes differently during the rapid period of brain wiring that happens in the first few years of life. Brain imaging studies show that young children with ASD, between ages 2 and 5, tend to have increased volume of both grey and white matter compared to typically developing children. This overgrowth pattern isn’t seen in older adolescents and adults with autism, suggesting a period of accelerated early brain growth followed by a slowdown or even volume loss later on.
At a finer level, the issue appears to involve how brain cells connect. Normally, the brain produces a surplus of synaptic connections in early childhood, then prunes away the ones that aren’t needed. Postmortem studies of autistic brains show a higher density of certain synaptic connections, particularly at excitatory synapses, along with connections that appear structurally immature. This pattern suggests that the pruning process doesn’t work as efficiently, leaving the brain with more connections but less organized wiring. Researchers also observe reduced long-range connectivity between the front and back of the brain, which may help explain why autism affects such a wide range of functions, from social processing to sensory integration.
Prenatal Environment and Maternal Health
What happens during pregnancy can shift autism risk, though the effects are much smaller than the genetic contribution. One of the clearest examples involves valproate, an anti-seizure medication. Children exposed to valproate during pregnancy had an absolute ASD risk of about 4.4 percent, roughly triple the rate among unexposed children. First-trimester exposure carried a particularly elevated risk. Other anti-seizure medications, like oxcarbazepine, did not show the same association, which tells researchers the risk is specific to certain drugs rather than to epilepsy itself.
Maternal infections during pregnancy also appear to matter. When a pregnant woman’s immune system mounts a strong response to an infection, the resulting inflammation can reach the developing fetal brain. Animal studies in mice, rats, and primates have confirmed that it’s the maternal immune response, not the specific virus or bacteria, that disrupts normal brain development. This doesn’t mean every cold or flu during pregnancy causes autism. Rather, a subset of pregnancies involving significant immune activation carry modestly increased risk.
Parental Age
Both maternal and paternal age are linked to higher autism risk. A large study tracking over 9,500 children diagnosed with ASD found that when either parent was 35 or older, the risk increased. The hazard ratios ranged from 1.21 to 1.65 depending on the combination of parental ages. For mothers under 35, the risk rose as the father’s age increased. For fathers under 35, the risk rose as the mother’s age increased. The effect is real but modest. Older parents are more likely to carry accumulated genetic mutations in their sperm or eggs, which may partly explain the connection.
Birth Complications in Preterm Infants
Babies born prematurely face a higher baseline risk for many developmental conditions, and autism is among them. Very low birth weight is one of the stronger signals: babies born under 750 grams had nearly four times the odds of an ASD diagnosis compared to babies born at a healthy weight. Babies between 750 and 1,499 grams had about double the risk. Once birth weight exceeded 1,500 grams, the association largely disappeared.
Brain hemorrhage during the newborn period also matters. Mild intracranial bleeding (grades 1 and 2) was associated with about 1.9 times the risk, while more severe bleeding (grades 3 and 4) carried 3.4 times the risk. Significant cerebral dysfunction in the newborn period, defined as seizures or abnormal neurological signs, was associated with roughly 4 to 5 times the risk. Interestingly, some factors that might seem risky, like birth asphyxia, respiratory distress syndrome, and neonatal jaundice, did not show consistent links to autism across studies.
Epigenetic Changes
Beyond the DNA sequence itself, the way genes are switched on and off also appears to differ in autism. These are called epigenetic changes. They don’t alter the genetic code but control whether specific genes are active or silent. In autism, researchers have found altered patterns of DNA methylation, the chemical tags that typically silence genes, across multiple brain regions. The cerebellum and temporal cortex of people with autism show reduced methylation on certain genes, meaning those genes may be inappropriately active.
Environmental exposures can trigger these epigenetic shifts. Certain chemicals increase DNA methylation patterns that have been associated with ASD. Changes in how tightly DNA is packaged around proteins called histones also play a role. Studies comparing brain tissue from people with autism to controls have found that regulatory elements are altered at a high rate. These findings suggest that autism risk isn’t just about which genes you carry but about how those genes are regulated during critical windows of development.
The Gut Microbiome Connection
Children with autism frequently experience gastrointestinal problems, and research has explored whether gut bacteria play a role. Gut microbes produce chemical messengers that can influence the brain through what’s called the gut-brain axis, communicating via the vagus nerve and through molecules that affect mood-related signaling systems. Children with ASD often show different gut bacteria profiles compared to neurotypical children, including higher levels of certain bacterial groups like Streptococcus and Ruminococcus.
However, the relationship is complicated. A microbiome study of 247 children found that the bacterial differences were largely a downstream consequence of dietary habits rather than a direct cause of autism. Children with ASD tend to eat a less varied diet, and less dietary diversity leads to less microbial diversity. The gut microbiome likely plays some role in symptom severity, particularly gastrointestinal discomfort, but current evidence doesn’t support it as a primary cause of autism itself.
Vaccines Do Not Cause Autism
A 2025 review by the World Health Organization’s expert committee on vaccine safety examined 31 studies published between 2010 and 2025, covering data from multiple countries, and found no causal link between vaccines and autism. This reaffirms the same conclusion the committee reached in 2002, 2004, and 2012. Vaccines containing thimerosal, aluminum, or any other standard ingredient do not increase autism risk. The original 1998 study that claimed a link was retracted due to fraudulent data, and its author lost his medical license. This is one of the most thoroughly investigated questions in pediatric medicine, and the answer is consistently clear.