The Genetics of Autism Spectrum Disorders

Joseph D. Buxbaum, PhD


December 29, 2005

In This Article


The distinction between "idiopathic" autism spectrum disorders (ASDs), and "secondary" ASDs, in which a known environmental agent, chromosome abnormality, or single-gene disorder can be identified, is an important one. Approximately 5%-10% of individuals with autism can be diagnosed with secondary autism; the remaining 90%-95% have idiopathic autism. In this review, the rare examples of ASDs resulting from environmental agents, including in utero exposures to rubella (German measles), valproic acid, and thalidomide, will not be discussed further.

To understand the causes of autism and ASDs, it is instructive to look back at careful studies completed as long as 30 years ago, before some of the current controversies about the causes of ASDs arose. Twin studies that are carried out in an epidemiologically appropriate manner can determine whether a disease results from environmental, genetic, or mixed causes. These studies involve ascertaining all individuals with ASDs in a particular region, which can involve a door-to-door search to identify all twins with at least 1 member of the twin pair affected with the disorder being studied.

Several now renowned researchers made some startling discoveries in the 1970s and '80s, when they conducted studies with no a priori hypothesis as to the cause of ASDs.[1,2,3,4] The most important outcome was the realization that the concordance rate for monozygotic (MZ) and dizygotic (DZ) twins differed substantially. Concordance rate is a measure of the probability that if 1 twin has an ASD, the other will have an ASD as well.

The important distinction between MZ and DZ twins arises because MZ twins are genetically identical, whereas DZ twins are genetically equivalent in the same way that non-twin siblings are. However, MZ and DZ twins, on average, will share the prenatal, perinatal, and postnatal environments equally with their co-twins. If MZ and DZ twins are equally concordant for a particular trait, that trait is very likely to be defined by nongenetic factors. As a trivial example, if one examines the concordance for the first language of MZ and DZ twins, if the first twin identified speaks English, the likelihood that the second twin also speaks English as a first language will not depend on whether the twins are MZ or DZ. The first language of the twins is entirely determined by nongenetic causes. If, however, the trait being studied is entirely genetic, then the MZ and DZ twins will have very different concordance rates.

If the trait being considered is a penetrant autosomal dominant genetic disorder (ie, where 1 copy of a mutation is sufficient to cause the disorder), it is essentially certain that if 1 MZ twin has the disorder, the other twin will have the disorder as well. However, in the case of DZ twins, if 1 twin has the disorder, there is a 50% chance that the other twin will be concordant, because the mutant gene may not be shared. In this case, the ratio of concordance between the MZ and DZ twins is not 1:1 (as it might have been with language) but rather 2:1.

In the case of a penetrant recessive genetic disorder (in which 2 copies of a mutation cause the disorder), if an MZ twin has the disorder, it is certain that the other twin will also have the disorder. Because this is a recessive disorder, however, if the twins are DZ, the probability that the second twin will have the disorder is 25%. In this case, the ratio of concordance between MZ and DZ twins is 4:1, a ratio indicative of a recessive disorder. As genetic disorders become more complex and involve more genes, the ratio of concordance between MZ and DZ twins becomes higher.

In the epidemiologic twin studies, the concordance ratio for autism in MZ compared with DZ twins was roughly 10:1. This finding furnished immediate and incontrovertible evidence that autism is a genetic disorder. These studies, carried out in the most rigorous manner, indicate that autism and also ASDs are genetic disorders.

Identifying autism and ASDs as genetic enabled subsequent studies to address several issues: the relationship between narrowly defined autism and other ASDs; the relationship between ASDs and other insults (such as perinatal complications); the genes underlying ASDs; and, ultimately, the diagnosis and treatment of ASDs.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.