Autism is not caused by any single thing. It develops through a combination of genetic factors and prenatal conditions, with most of the brain changes beginning before a child is born. Large twin studies estimate that genetics account for 64% to 91% of autism risk, making it one of the most heritable neurodevelopmental conditions. The rest of the risk comes from environmental factors, nearly all of which act during pregnancy rather than after birth.
Genetics Play the Largest Role
The strongest evidence for a genetic basis comes from twin studies. Identical twins, who share virtually all their DNA, have a concordance rate of about 70% for autism. That means if one identical twin is autistic, there’s roughly a 7 in 10 chance the other twin is too. Fraternal twins, who share about half their DNA like any siblings, show a concordance rate near 0% in population-based studies. That enormous gap points squarely to genes.
No single “autism gene” exists. Researchers have identified hundreds of genes that each contribute a small amount of risk. Some of these genes are active early in pregnancy, affecting how brain cells multiply, migrate, and organize into layers. Others become active later in pregnancy and in the first months of life, shaping how brain cells form connections and wire into networks. In most cases, a child inherits a combination of common genetic variants from both parents that together push risk above a threshold. Less commonly, a new genetic mutation arises spontaneously in the child rather than being inherited.
Sibling studies reinforce this picture. If one child in a family has autism, the chance of a younger sibling also being diagnosed is roughly 2% to 6%, which is several times higher than the general population rate. The risk is higher still when the older sibling’s autism is more severe or when both parents carry relevant genetic variants.
Brain Development Begins Changing Before Birth
Autism is not something that appears suddenly in toddlerhood, even though that’s when parents typically notice behavioral differences. Postmortem brain studies of autistic children as young as two have found clear signs of atypical development that trace back to the first and second trimesters of pregnancy. During weeks 10 to 20 of gestation, the brain produces cortical neurons at an exponential rate. In some autistic individuals, researchers find an excess of neurons in certain brain regions, and because those neurons are only generated prenatally, the difference had to originate during pregnancy.
The developmental disruptions unfold in two broad waves. In the first and second trimesters, genes involved in cell growth, brain cell production, and cell migration are affected. In the third trimester and early postnatal life, a different set of genes disrupts how neurons extend branches, form connections, and organize into functional circuits. About two-thirds of the high-confidence autism risk genes act during that earlier wave, with the remaining third acting in the later one.
Prenatal Environment and Maternal Health
While genes set the foundation, conditions during pregnancy can raise or lower risk. Three maternal health factors have the strongest evidence: infection during pregnancy, gestational diabetes, and obesity. Each of these creates an inflammatory environment that can affect fetal brain development. Obesity increases inflammatory signaling molecules in the mother’s bloodstream, and in animal studies, this maternal inflammation leads to inflammation in the offspring’s developing brain. Diabetes produces a similar inflammatory state in uterine tissues and also causes high blood sugar, which can trigger a chain reaction in the fetus: the baby overproduces insulin to compensate, burns through more oxygen, and experiences chronic low oxygen levels and oxidative stress in the womb.
When obesity and diabetes occur together, the risk compounds. One large study found that children born to mothers who were both obese and had gestational diabetes were about three times more likely to be diagnosed with autism compared to children of mothers with neither condition. When the mother had pre-existing diabetes combined with obesity, the risk was nearly four times higher. For autism occurring alongside intellectual disability, those combined risks were even more dramatic, reaching six to nearly ten times higher.
Other prenatal exposures under active investigation include certain antidepressants, antibiotic use, and environmental toxicants, though the evidence for these is less established than for diabetes and obesity.
Parental Age at Conception
Both maternal and paternal age independently affect autism risk. For every 10-year increase in a mother’s age, the risk rises by about 31%. For fathers, each additional decade raises risk by about 28%. To put that in concrete terms: the cumulative chance of an autism diagnosis by age 10 nearly doubled when comparing the youngest mothers (under 20) to the oldest (40 and over). For fathers, the difference was even starker, with more than a threefold increase from the youngest to the oldest group, independent of the mother’s age.
The biological explanation likely involves the accumulation of new mutations in sperm over time and age-related changes in egg quality. Older parents are more likely to pass along spontaneous genetic mutations that weren’t present in their own genomes.
Premature Birth and Low Birth Weight
Babies born very early or very small face elevated autism risk. Infants born before 32 weeks of gestation are about two and a half times more likely to be diagnosed with autism compared to full-term babies. Very low birth weight babies (under about 3.3 pounds) have more than three times the odds of an autism diagnosis compared to babies in the normal weight range. Even moderately low birth weight (under about 5.5 pounds) raises the odds by roughly 60%.
Babies who are small for their gestational age, meaning they weigh less than expected for how far along the pregnancy was, carry about 70% higher odds. These associations held up even after researchers accounted for factors like maternal age, smoking, psychiatric history, and congenital anomalies. Notably, these birth-related risk factors were linked to autism with intellectual disability but not to Asperger syndrome specifically, suggesting they may affect certain developmental pathways more than others.
Vaccines Do Not Cause Autism
This is one of the most thoroughly studied questions in modern medicine, and the answer is unequivocal. Multiple case-control studies conducted in the United Kingdom, United States, Poland, and Japan have found no increased risk of autism following MMR vaccination. The largest of these, which included over 1,200 cases and 4,400 controls, found a relative risk of 0.86, meaning vaccinated children were actually slightly less likely to develop autism than unvaccinated children (though the difference was not statistically significant).
A Danish cohort study tracked more than 500,000 children, including nearly 100,000 who were unvaccinated, and found no association between the MMR vaccine and autism. A study of about 100,000 younger siblings of children already diagnosed with autism, a group that would theoretically be most vulnerable, found no increased risk from vaccination even in this high-risk population. Three separate reviews by the National Academy of Medicine have all reached the same conclusion. A meta-analysis pooling the published evidence confirmed it. The original 1998 study that sparked the concern was retracted due to fraud, and its author lost his medical license.
How Common Autism Is Today
As of the most recent CDC surveillance data from 2022, about 1 in 31 eight-year-olds in the United States has been identified with autism. That’s up from 1 in 150 in the year 2000 and 1 in 36 in 2020. The increase is largely attributed to broader diagnostic criteria, greater awareness, and better identification in groups that were historically underdiagnosed, including girls, Black children, and Hispanic children.
Autism remains 3.4 times more common in boys than girls. Prevalence varies substantially by location, ranging from about 1 in 100 in some areas to 1 in 19 in others, reflecting differences in screening and diagnostic access rather than true differences in how often autism occurs. Children born in 2018 were being identified with autism by age four at 1.7 times the rate of children born just four years earlier, suggesting that earlier diagnosis is becoming more common rather than autism itself becoming dramatically more frequent.
How These Factors Work Together
Autism develops when genetic susceptibility meets certain prenatal conditions during critical windows of brain development. A child might carry dozens of gene variants that each nudge brain development in a slightly different direction. If that genetic load is high enough on its own, autism develops regardless of other factors. If the genetic load is moderate, prenatal conditions like maternal inflammation, low oxygen, or oxidative stress may tip the balance. This is why autism runs in families but doesn’t follow a simple inheritance pattern, and why two children with the same parents can have very different outcomes.
There is also a layer of gene regulation at work. Chemical tags on DNA and on the proteins that package it can switch genes on or off without changing the genetic code itself. These modifications are influenced by the prenatal environment, meaning that factors like maternal health and toxic exposures can alter how autism-related genes behave. This helps explain how environmental risk factors translate into changes in brain development even when a child’s underlying DNA sequence looks unremarkable.