To a large extent, the development of autism is associated with genes, but the genetics of autism are complex, and it is unclear whether the prevailing influence on the emergence of autism spectrum disorders is the interaction of many genes or rare mutations with a strong effect. The complexity is due to the multifaceted interactions of a large number of genes, the external environment, and epigenetic factors, which in themselves do not change the DNA code, but can be inherited and modify gene expression.

In early twin studies, the heritability of autism was estimated at more than 90%, provided children live in the same environment and the absence of other genetic and medical syndromes. However, most of the mutations that increase the risk of autism remain unknown. As a rule, in the case of autism, it is not possible to trace the connection of the disorder to a Mendelian mutation (affecting a single gene) or to a single chromosomal aberration, as in Angelman or Martin-Bell syndrome. A number of genetic syndromes have been associated with autism spectrum disorders, but in none of them do the symptoms fit exactly as typical of such disorders. Many candidate genes have been found, but the effect of each of them is very small.

Variations in the copy number – spontaneous deletions and duplications of genomic regions in meiosis – can cause a large number of autistic people in healthy families. Consequently, a significant number of cases can be attributed to genetic changes that are highly inherited, but, moreover, were not inherited themselves: these are new mutations that caused autism in the child, but were absent in the parents.

Experiments with gene replacement in mice suggest that autism symptoms are closely related to late developmental stages, in which synaptic activity and related changes play an important role, and that replacing genes or modulating their activity after birth can alleviate symptoms or reverse reversal of violations. All known teratogens (substances that cause birth defects) associated with the risk of autism are reported to have an effect within the first eight weeks after conception. While these findings do not exclude the possibility of later triggering or influencing the mechanisms of autism, they provide strong evidence that the disorder lies in its earliest stages of development.

There is only fragmentary data on other external factors that can cause autism, and they are not confirmed by reliable sources, but an active search is underway in this direction. Numerous environmental factors have been reported about possible contributions to autism or the aggravation of the disorder, and some of the alleged interactions may be useful as research subjects. These factors include certain foods, infectious diseases, heavy metals, solvents, diesel exhaust, PCBs, phthalates and phenols used in plastics, pesticides, brominated flame retardants, alcohol, smoking, drugs and prenatal stress.

The time when a child is routinely vaccinated may coincide with the time the parents first notice their autistic symptoms. Concerns over the role of vaccines have led to declining immunization rates in some countries, increasing the risk of measles outbreaks. However, the vast majority of scientific studies have not found a link between the MMR vaccine and autism, as well as convincing scientific evidence of the effect of thimerosal added to vaccines on the risk of developing autism.

Some researchers associate the occurrence of autism with a malfunction (hyperactivity) of the patient’s immune system.