Autism: Environmental and Genetic? Clues from parental age.

By Nestor Lopez-Duran PhD

The last issue of the prestigious Journal of Child Psychology and Psychiatry included a study that examined the role of parental age on the risk for autism among two very large nationally representative cohorts in two separate countries (Sweden and the UK). The basic question whether the parents age impact the risk of having a child with autism is not that exciting or innovative as several studies have shown that older parents, especially dads, are more likely to have children with autism. What was really interesting about this study is that it was conducted with an incredibly large numbers of twins, which can help us understand the association between parental age and the relative environmental vs. genetic contributions to autism.

Before I talk about the study I want to provide some background on autism research and twin studies. The question that most behavioral geneticists ask is NOT whether autism is genetic or environmental. There is enough data to show that autism is NOT purely genetic and that autism is NOT purely environmental. The consensus is that autism is a very heterogenous condition that is likely due to multiple genetic and environmental factors.  So real question is what are the relative contributions of the environment, our genes, and other bio-social processes to the development of autism. To this end, behavioral geneticists examine the similarity between monozygotic (MZ) vs. dizygotic (DZ) twins to determine the relative genetic vs. environmental contributions of a specific condition. Specifically, if the correlation within MZ twins in regards to the rate of a disorder is greater than the correlation within DZ twins, then you would assume a significant genetic contribution. Why? MZ twins are genetically identical while DZ are not.  If a disorder has a large genetic contribution, then you would expect those twins that are identical to be more likely to both have the disorder than twins that are not identical. In contrast, in a disorder with little genetic contribution, DZ and MZ twins would be equally likely to share the disorder since the difference in how genetically identical they are would make little difference.

So in this study, the authors examined data from two large twin cohorts from Sweden (N=11,122) and the UK (N= 13,524) who were assessed at age 9 with two different autism scales/interviews. In Sweden the children were assessed with the Autism-Tics, AD/HD, and other Co-morbidities (A-TAC). In the UK, the children were assessed with the Childhood Autism Spectrum Test.

The results:

The graphic below shows the percentage changes in the probability of having a diagnosed ASD by having a father in different age groups.

As you can see, compared to 24-34 year old dads, there was a large increase in the odds of having a child with ASD (almost 100%) for younger dads, a similar increase for dads 35-44, and a very large increase (over 200%) for dads older than 51. However, only the change in fathers >51 in Sweden was statistically significant. The other changes only approached significance, likely because of the low rates of ASD among these cohorts.

As comparison, below you can see the changes in ASD traits for children of fathers in different age groups.

This time, the increase in autism traits for children of young fathers (<25) and older (>50) fathers is statistically significantly when compared to kids whose fathers were 25-34. So this study is consistent with previous research showing an increased risk of ASD among older fathers. However, the study also shows an potential increased risk of ASD for younger fathers as well. There was no effect of maternal age on the risk of ASD.

What about the role of parental age in the relative genetic/environmental contribution to ASD diagnoses?

Below is a graph that presents the correlation within MZ or DZ twins for different paternal age groups.

You will notice that the correlation between the MZ twins is always higher than the correlation between the DZ twins, suggesting some genetic contribution to the disorder. That is, MZ are more likely to BOTH have ASD than DZ twins. However, notice how the difference between the DZ and MZ twins is reduced for the older parents, in both Sweden and the UK. What does this mean? It means that the relative genetic contributions to ASD appear to decrease for older fathers. Now see below the raw correlations for all age groups:

What it is interesting about these data is that the correlation within the MZ increases with the fathers age. For example,  MZ twins of fathers over 40 have an almost 1-to-1 correspondence of the disorder. That is, if one twin had the condition, the other twin almost always had it too. Does this means genetic? Well, at the surface you would think this means genetic, after all both twins are genetically identical and both twins have the disorder. However, remember that MZ were conceived from the same sperm, and in this case, from the same sperm that may be damaged. So the increase concordance among MZ twins for older dads is not necessarily reflective of a genetic anomaly. In fact, the authors indicated how this effect may be due to the prolonged exposure to environmental toxins among the older fathers leading to sperm mutations. If that hypothesis is correct, it could be the environment, and not the genes, what is responsible for the increase risk in ASD among children of older dads.

The reference:
Lundström, S., Haworth, C., Carlström, E., Gillberg, C., Mill, J., Råstam, M., Hultman, C., Ronald, A., Anckarsäter, H., Plomin, R., Lichtenstein, P., & Reichenberg, A. (2010). Trajectories leading to autism spectrum disorders are affected by paternal age: findings from two nationally representative twin studies Journal of Child Psychology and Psychiatry DOI: 10.1111/j.1469-7610.2010.02223.x