Brief Report: On the Concordance Percentages for Autistic Spectrum Disorder of Twins

Brief Report


In the development of genetic theories of Autistic Spectrum Disorder (ASD) various characteristics of monozygotic (MZ) and dizygotic (DZ) twins are often considered. This paper sets forth a possible refinement in the interpretation of the MZ twin concordance percentages for ASD underlying such genetic theories, and, drawing the consequences from that refinement, a possible early environmental factor in the later development of ASD.


Autistic spectrum disorder Concordances Twins 

In the development of genetic theories of Autistic Spectrum Disorder (ASD) various characteristics of monozygotic (MZ) and dizygotic (DZ) twins are often considered. The purpose of this paper is to set forth a possible refinement in the interpretation of the MZ twin concordance percentages for ASD underlying such genetic theories, and, drawing the consequences from that refinement, suggest an early environmental factor in the later development of ASD.

In their paper entitled “A unified genetic theory for sporadic and inherited autism”, Zhao et al. (2007) state in the first paragraph of discussion of their results: “We must emphasize at the outset that our biological interpretation of the risk models assumes that risk is determined by genetic factors, and thus, except for their appealing simplicity, the risk models themselves should not be taken as evidence for genetic causation. We cannot rule out environmental factors, such as complications during pregnancy, contributing to the observed risk data, and the presence of these factors could impact our biological interpretation and some of our modeling assumptions. Nevertheless, in what follows, we assume a genetic basis for risk, justified largely by MZ and DZ twin studies.”

The present authors added the italics to the quoted paragraph to indicate matters addressed below. The literature provides concordance percentages for ASD of monozygotic (MZ) twins ranging from about 60% to as high as 90%, and for dizygotic (DZ) twins from about 5–10% up to 20+%. Concordance rates for siblings generally are given in the range of ~3–6% (Bailey et al. 1995; Sebat et al. 2007; Rutter 2005a, b). The large spread in the percentages quoted for MZ concordance may be due, at least in part, to the small numbers in the MZ twin databases available for ASD studies. However, here we propose a more fundamental basis for the MZ concordance rate spread.

Specifically, the spread of percentages reported may come about because the MZ twin databases from which they are derived do not differentiate between monzygotic-monochorionic (MZ-MC) and monozygotic-dichorionic (MZ-DC) pairs of twins. We are not aware of the existence of any databases for ASD concurrence rates that do distinguish between MZ-MC and MZ-DC pairs. Note that the ratio of MZ-MC/MZ-DC twin births is ~2:1 (Benirschke et al. 2006; Gilbert-Barness and Debich-Spicer 2004). Although it may be coincidental, this ratio is approximately the same as the frequently cited 60–70% concordance rates for MZ twins. Evidence for a large difference between MC and DC twin concurrence rates is presented below.

MZ twin zygotes are genetically identical (or very close to identical) shortly after division. They then develop into two embryos either within one chorion, although almost always in separate amniotic sacks (MZ-MC-DA), or with a relatively impermeable, thick (at least in the first trimester) chorionic membrane between the two amniotic sacks (MZ-DC-DA). [MZ-MC-MA twins are very unusual—less than 1%, (Blickstein 2006)]. It is known that amniotic fluids from neighboring amniotic sacks penetrate fairly readily through the amniotic membranes—but the thicker, multilayered chorionic membranes between amniotic sacks impede such fluid interchanges. Thus, the uterine environments of a pair of MZ-MC twin zygotes/embryos/fetuses are much closer to being identical than the uterine environments of a pair of MZ-DC twin zygotes/embryos/fetuses. Indeed, after division the MZ-DC twin zygotes’ uterine environments are quite similar to DZ twin zygotes’ uterine environments—at least those DZ twin zygotes that develop into same sex twin fetuses. Therefore it is reasonable to suggest that the ASD concordance rate of MZ-DC twins may be close to that of DZ (same sex) twins, and very different from that of MZ-MC twins.

Equation 1 is a simple mathematical representation of this hypothesis
$$ fR_{ 1} + \left( {1 - f} \right)R_{ 2} = R_{ 3} , $$
where f is the ratio of MZ-MC twin fetuses to all MZ fetuses, and thus (1−f) is the ratio of MZ-DC twin fetuses to all MZ fetuses; R1 is the concordance rate for MZ-MC twins; R2 is the concordance rate for MZ-DC twins, and R3 is the concordance rate for MZ twins. Figure 1 gives a graphic presentation of Eq. 1 with f = 2/3. Note that for R1 ≅ 0.95 and R3 ≅ 0.65, the R2 rate is within the commonly quoted DZ rates.
Although we recognize that the genes suspected of involvement with ASD are presumably not the same as those connected to Down’s syndrome, we do note that fetal ultrasonic and other data cited in Matias, Montenegro and Blickstein (2006), indicate that there is a very substantial difference in Down’s concordance percentages between MZ-MC and DC (including both MZ and DZ) twin fetuses. Since Down’s syndrome can be detected with a high probability in the uterus, compelling reasons for identifying whether at risk fetuses are mono- or dichorionic have existed for Down’s which to date have not existed for ASD.
Fig. 1

A plot of R2 = 3R3−2R1 (i.e. Eq. 1 with f = 2/3) for three different values of R1, where R3 is the concordance rate all MZ twins, R1 the rate for MZ-MC twins, and R2 the rate for MZ-DC twins


Though differing theories of the origins of ASD are extant, there is general agreement that ASD is a genetic disorder. If the MZ-MC and MZ-DC twin concordances for ASD do indeed differ substantially as outlined above, it would follow that two zygotes/embryos/fetuses, originating from the splitting of one fertilized ova, developing in as nearly identical uterine environments as are available (MC), have about a 95–100% concordance, while two other zygotes, also originating from the splitting of one fertilized ova, developing in similar but not identical uterine environments (DC) have concordance rates much nearer those of DZ twins and of ordinary siblings. The differences between MZ-MC and MZ-DC environments are likely to be most pronounced in the first trimester because the chorionic membrane separating the two DC fetuses becomes less dense in the 2nd and 3rd trimester. Clearly, the separately developing placentas of DC twins are also pertinent. These circumstances suggest that a uterine environmental factor or trigger, present or developed in the first trimester of pregnancy, is necessary for the zygote/embryo/fetus development eventually resulting in the post-natal diagnosis of ASD. Rutter’s comment (2005a, p. 426, para 4) that “An examination of 16 MZ pairs concordant for autism or ASD showed that there was enormous clinical heterogeneity even when pairs shared exactly the same segregating genetic alleles” is, at a minimum, consistent with the above reasoning. Pushing further with this speculation, presumably the amniotic fluids in which MZ-MC-DA twin zygotes/embryos/fetuses are immersed are much closer to being identical in composition and in the concentrations of various components than are the amniotic fluids of MZ-DC-DA twin zygotes/embryos/fetuses—or same sex (let alone opposite sex) DZ zygotes/embryos/fetuses. Bohm et. al. (2007) have suggested that measuring the testosterone concentration in amniotic fluid samples (TECAFs), particularly for male fetuses, might provide useful information. In any case, we believe that preliminary support or contradiction of the above conjectures would not be a difficult or expensive undertaking for a clinical research group. A possible approach to such an effort would be to look at the differences between same sex twin TECAFs. We would expect that the average TECAF difference between twin fetuses would be smaller for MZ-MC pairs than for MZ-DC or DZ pairs. Lastly, if the MZ-MC ASD concordance rate does prove to be very different from the MZ-DC rate, it would suggest that future twin studies for many different purposes should consider whether MZ-MC/MZ-DC differences might be important.



It is a pleasure to acknowledge helpful discussions with Dr. Dan Rappolee, School of Medicine, Drs. Merlin Ekstrom and Karen Rossman, Division of Laboratory Animal Resources, Wayne State University, and Drs. David Loeffler and Dianne Camp of the Beaumont Hospital Neuroscience Laboratory, Royal Oak, MI. The journal searches by Ms. Caralee Witteveen-Lane, then a student in the WSU Library and Information Sciences Program, were useful in the preparation of this manuscript.


  1. Bailey, A., LeCouteur, A., Gottesman, I., Bolton, P., Simonoff, E., Yuzda, E., et al. (1995). Autism as a strongly genetic disorder: Evidence from a British twin study. Psychological Medicine, 25, 63–77.PubMedCrossRefGoogle Scholar
  2. Benirschke, K., Kaufmann, P., & Baergen, R. (2006). Pathology of the Human Placenta (5th Edition), Springer Science+Business Media Inc., New York 877–885. (Note that the 2:1 ratio for MZ-MC/MZ-DC is for the US population, and is different for some other ethnic/racial groups).Google Scholar
  3. Blickstein, I. (2006). Monozygosity. In A. Kurjak & F. A. Chervenak (Eds.), Textbook of Perinatal Medicine, v2, (2nd edition) Chap. 146, Informa UK Ltd. Boca Raton, FL: Taylor & Francis.Google Scholar
  4. Bohm, H. V., Fry McComish, J. E., & Stewart, M. G. (2007). On a possible early identification procedure for babies at high risk for autistic spectrum disorder. Medical hypotheses, 69, 47–51. doi:10.1016/j.mehy.2006.09.073.PubMedCrossRefGoogle Scholar
  5. Gilbert-Barness, E., & Debich-Spicer, D. (2004). Embryo and Fetal Pathology. Chap. 23, (p. 622). Cambridge, UK, New York: Cambridge University Press.Google Scholar
  6. Matias, A., Montenegro, N., & Blickstein, I. (2006). Sonographic evaluation of multiple pregnancies. In A. Kurjak & F. A. Chervenak(Eds.), Textbook of Perinatal Medicine, v2, (2nd edition), Chap. 150, Informa UK Ltd. Boca Raton, FL: Taylor & Francis.Google Scholar
  7. Rutter, M. (2005a). Genetic Influences and Autism. In F. R. Volkmar (Ed.), Handbook of Autism and Pervasive Developmental Disorders, v1: Diagnosis, Development, Neurobiology, and Behavior (3rd ed., pp. 425–452). Hoboken NJ: Wiley.Google Scholar
  8. Rutter, M. (2005b). Aetiology of autism: Findings and questions. Journal of Intellectual Disability Research, 49, 231–238. doi:10.1111/j.1365-2788.2005.00676.x.PubMedCrossRefGoogle Scholar
  9. Sebat, J., Lakshmi, B., Malhotra, D., Troge, J., Lese-Martin, C., Walsh, T., et al. (2007). Strong association of De Novo Copy number mutations with autism. Science, 316, 445–449. doi:10.1126/science.1138659.PubMedCrossRefGoogle Scholar
  10. Zhao, X., Leotta, A., Kustanovich, V., Lajonchere, C., Geschwind, D. H., Law, K., et al. (2007). A unified genetic theory for sporadic and inherited autism. Proceedings of the National Academy of Sciences of the United States of America, 104, 12831–12836. doi:10.1073/pnas.0705803104.PubMedCrossRefGoogle Scholar

Copyright information

© Springer Science+Business Media, LLC 2009

Authors and Affiliations

  1. 1.Department of Physics and AstronomyWayne State UniversityDetroitUSA

Personalised recommendations