You are here:
-A A +A

Twins! A Key to the Mysteries of Autism

Date Last Revised: 
August 21, 2014
Date Published: 
October 6, 2009

Photo of twin boys illustrating autism twins researchResearchers from the IAN Project, along with other researchers, have published an important paper in the Archives of Pediatrics & Adolescent Medicine entitled "Characteristics and Concordance of Autism Spectrum Disorders Among 277 Twin Pairs."1 It is based on information provided by the families of twins to IAN Research.

Why is twin research important? Below, we discuss the importance of twin research and IAN’s findings.

Families of children on the autism spectrum make valuable contributions to research. This is true whether their children are younger or older, and no matter their diagnosis, gender, level of functioning, or how many siblings they have. One type of family is very important to researchers, however: families of twins.

The Importance of Twins in Autism Research

Twins are especially important in autism research because studying them can help researchers untangle genetic and environmental factors in autism.

Twins offer researchers a rare opportunity to compare people who share all of their genes (identical twins) and people who share only half of their genes (fraternal or nonidentical twins). Twins also offer a chance to compare two individuals when environmental factors were likely the same for both. Although all full siblings share 50% of their genes, those that were born at different times may have been exposed to different environments, while twins -- even nonidentical twins -- are likely to have shared a very similar environment. For example, they shared a womb, so if their mother was sick or stressed during pregnancy, or took certain medications, they were probably both affected in the same way. They were probably fed the same foods (with the same nutrients, preservatives, etc.), lived in the same home (near the same factory), and breathed the same air (pollution and all).

Since twins likely shared the same environment, researchers can be fairly confident that differences they observe in twins pairs are due to genetic factors. They begin by looking at twin pairs where at least one twin has an autism spectrum disorder (ASD). Then they split the twin sets into two groups: those that are identical and those that are not. The researchers have special terminology for this. Since identical twins came from a single fertilized egg, or zygote, they are called monozygotic, or MZ. Nonidentical or fraternal twins came from two separate fertilized eggs, so they are called dizygotic, or DZ.

The next step is to see how often both twins have an ASD. The word researchers use for this is concordance. Basically, if a twin set is concordant for autism, both twins have autism. If it is discordant, one twin has the condition, and one does not. Depending on their aims, researchers may be very strict and consider only twins who have the exact same ASD diagnosis concordant, or they may broaden the definition and consider twins concordant as long as both have any kind of ASD, whether it’s autism, Asperger syndrome, or pervasive developmental disorder-not otherwise specified (PDD-NOS).

Researchers then calculate a percentage to represent concordance:

Concordance % = Number of Concordant Twin Pairs ÷ Total Number of Twin Pairs

Twin studies up to this point have shown that identical twins with autism have been more than 90% concordant, while nonidentical twins have been less than 31% concordant.2 3 4  This is one of the major reasons researchers have concluded there is a large genetic contribution to autism. In brief, twins who share more genes are more likely to share autism, too.

The Contribution of IAN Families

One difficulty faced by twin studies in autism has been, simply put, finding twins. The most well-known studies to date have involved fewer than 50 twin sets. IAN, in contrast, now has 277 twin sets -- more than researchers have ever been able to study at one time. (See Table 1.)

This permits researchers to examine factors they couldn’t possibly explore when participating twins were so few in number, such as differences by ASD diagnosis and differences by gender. Furthermore, because IAN families have shared so much information about their children’s development and history, researchers can explore additional issues across twin sets, like regression or intellectual disability.

What do the IAN twin sets show? With this larger sample, are the percentages for concordance in identical and nonidentical twins similar to what researchers have found in the past?

The answer is yes. Analysis of the IAN data results in concordance rates for identical and nonidentical twins that are very similar to the findings of past researchers. Our 88.1% concordance for identical twins is very close to the 91 to 95% concordance found in earlier studies. Our rate of 30.5% concordance for nonidentical twins is comparable to the 30.8% rate found in a recent Japanese study that, like IAN, looked at children with not just autism, but also PDD-NOS and Asperger syndrome. (See Table 1.)

Table 1. Results of Several Twin Studies

  Total Twin Sets Phenotype
Twins concordant if both have:
(% of twin sets where both
have ASD)
(% of twin sets
where both
have ASD)
UCLA, 19852 40 Autism only 22 out of 23 pairs (95.7) 4 out of 17 pairs (23.5)
University of Goteburg, Sweden, 19893 21 Autism only 10 out of 11 pairs (91) 0 out of 10 pairs (0)
Nagoya Child Welfare Center, Japan, 20084 45 Autism, PDD-NOS, or Asperger’s 18 out of 19 pairs (94.7) 8 out of 26 pairs (30.8)
IAN Sample, 20091 277 Autism, PDD-NOS, Asperger’s, or Other ASD 59 out of 67 pairs (88.1) 64 out of 210 pairs (30.5)

This is a very important confirmation of earlier findings in a much larger sample made possible thanks to IAN families of twins.

Other Findings Based on the IAN Twin Data

Because IAN collects a great deal of information from families on a wide range of issues, the IAN twin study has been able to compare twins in new ways.

Timing of Parent’s First Concern

Parents of nonidentical twins were more likely than parents of identical twins to become concerned about their children’s development before 18 months of age. This may be because nonidentical twins did not both have an ASD as often as identical twins did. It may be that parents who had one typically developing twin could more easily see that the affected twin was lagging behind or behaving differently. Parents with two twins on the spectrum would not observe this contrast, and so might not become concerned as soon.

Differences in Development, Regression, or Cognitive Impairment More Likely in Nonidentical Twins Where Both Have ASD

As you might expect, nonidentical twin pairs in which both children had an ASD tended to be less alike than concordant identical twin pairs in a variety of ways. They were more likely to differ in intellectual ability, the time when they reached developmental milestones (like walking or saying words), or whether they experienced any kind of skill loss or regression.

Asperger Syndrome: Inherited Differently?

Autism and PDD-NOS seemed to be inherited in a similar way, but Asperger syndrome seemed to be inherited differently. In identical twin sets where both children had some form of ASD, it was very unusual for one to have Asperger’s and the other to have autism or PDD-NOS. In other words, when one identical twin had Asperger’s, the other usually, did, too.

How Does Autism 'Miss' One Twin?

In identical twin pairs, there are cases where one twin does not develop an ASD and one does. What accounts for this? How did autism “miss” the unaffected twin?

This is an area where many theorize that environmental factors have come into play. Was the affected twin exposed to some environmental hazard? Possible environmental culprits are being explored by a number of researchers, as they investigate everything from air pollution to pesticides. Did some exposure turn a certain gene “on” or “off,” or have some other effect?

(For articles on research on possible environmental factors in autism, see Environmental Aspects of Autism.)

Another possibility is a difference in the birthing process. Did one experience some kind of trauma, such as oxygen deprivation, before, during, or after birth, that the other did not?

One fascinating idea being explored is that sometimes identical twins may be different because they did not share a chorion, which can be thought of as a sac beyond and surrounding the amniotic sac. Twins can have their own chorion, or share one. If identical twins did not share a chorion, then they have actually not shared the exact same environment in the womb. Will different ASD concordance rates be found for identical twins who did or did not share a chorion?5

As researchers continue to study simplex families (that have only one child with ASD), multiplex families (that have more than one child with ASD), and both identical and nonidentical twin sets, the complex genetic and environmental picture in autism should become much more clear.

Families: Researchers Continue to Need Your Help

To continue to make progress, researchers need the participation of families. We thank each and every family that is participating in the IAN Project. Please do continue to fill in your questionnaires, as every bit of data is important.

To parents of twins -- a special plea: As you can see from this discussion, information about your twins and your family is especially critical. Please fill in all your IAN questionnaires! With twins, it’s even more challenging to find time, but remember that you can fill in our online questionnaires at any time of the day or night, doing just as much as you are able, and continuing where you left off when you get another few minutes free.

All families: We are here to help! Let us know if you have any questions or concerns at or 866-348-3440.

Please rate the helpfulness of this article: 
Average: 4.8 (10 votes)
Additional Resources: 
  1. Rosenberg, R.E., Law, J.K., Yenokyan, G., McGready, J., Kaufmann, W.E., & Law, P.A. (2009). Characteristics and Concordance of Autism Spectrum Disorders Among 277 Twin Pairs. Archives of Pediatric and Adolescent Medicine, 163(10):907-914. View abstract.
  2. Ritvo, E. R., Freeman, B. J., Mason-Brothers, A., Mo, A., & Ritvo, A. M. (1985). Concordance for the syndrome of autism in 40 pairs of afflicted twins. The American Journal of Psychiatry, 142(1), 74-77. View Abstract
  3. Steffenburg, S., Gillberg, C., Hellgren, L., Andersson, L., Gillberg, I. C., Jakobsson, G., et al. (1989). A twin study of autism in Denmark, Finland, Iceland, Norway and Sweden. Journal of child psychology and psychiatry, and allied disciplines, 30(3), 405-416. View Abstract
  4. Taniai, H., Nishiyama, T., Miyachi, T., Imaeda, M., & Sumi, S. (2008). Genetic influences on the broad spectrum of autism: Study of proband-ascertained twins. American Journal of Medical Genetics. Part B, Neuropsychiatric Genetics: The Official Publication of the International Society of Psychiatric Genetics, 147B(6), 844-849. View Abstract
  5. Bohm, H. V., & Stewart, M. G. (2009). Brief report: On the concordance percentages for autistic spectrum disorder of twins. Journal of Autism and Developmental Disorders, 39(5), 806-808. View Abstract 
  6. Geschwind, D. H., Sowinski, J., Lord, C., Iversen, P., Shestack, J., Jones, P., et al. (2001). The autism genetic resource exchange: A resource for the study of autism and related neuropsychiatric conditions. American Journal of Human Genetics, 69(2), 463-466. View Abstract