To what extent are correlations of father's age with birth defects and autism causal?

Question

I read the following in the paper today:

Dr Karin Hammarberg ... said that while most children are born healthy, large studies of parental age were starting to show higher rates of birth defects and autism in children born to men over 40.

A recent review of paternal age published in the Asian Journal of Andrology said an American study of 132,000 men found children of those over 45 were nearly six times more likely to have an autism spectrum disorder compared to children born to men under 30.

The review also pointed to a Dutch study of 60,000 births which found children born to men over 40 were three times more likely to have autism and a US study of 5 million births which showed men over 50 had a 15 per cent higher chance of having a baby with birth defects including congenital heart disease and cleft palates.

However, my first thought was that the age of the father must be highly correlated with the age of the mother. Thus, an alternative explanation is that the higher prevalence of birth defects for older fathers is due to the fact that older fathers typically have babies with older mothers. Of course, the same argument could be made for the reported higher prevalence of birth defects in older mothers (i.e., they're having babies with older fathers).

Questions:

• To what extent is paternal age associated with prevalence of birth defects after controlling for maternal age?
• To what extent is any association between paternal age and prevalence of birth defects causal? (maternal age is the main covariate that I could think of, but there might be other important confounding variables)

Answer 1

This is a partial answer suggesting a possible causal mechanism.

One of the factors leading to birth defects and autism is mutations in the parent's genetic material. In a recently published article, Sun et al. (2012) observed that:

The paternal-to-maternal mutation rate ratio is 3.3, and the rate in fathers doubles from age 20 to 58, whereas there is no association with age in mothers.

In other words, while the mother's age will not have an effect on the number of mutations passed to the offspring, the older the father the more mutations. That means a possible causal mechanism for correlations of father's old age is an increase in the number of mutations (most of which are neutral or detrimental) passed to the offspring.

Kong et al. (2012) explain the connection of de novo mutations (of the sort observed by Sun et al. (2012)) to autism and other disorders more clearly. This blog post suggests why fathers might produce more mutations:

The difference probably comes from the distinct biology of male and female reproduction. Cells that give rise to sperm cells continuously divide throughout a man’s life while cells that develop into ovules stop replicating when the woman is still a fetus. Consequently, an ovule at any time of a woman’s life will have gone through 24 cell divisions while sperm will be at about 200 divisions in a 20 year old man and up to 600 in a man of 40 years. When cells divide, they replicate their DNA to transmit a copy of the genome to each daughter cell but this replication is not 100% accurate. With each division mutations creep in the DNA so the more divisions a cell goes through, the more mutations it accumulates.

References

Kong, A., Frigge, M.L., Masson, G., Besenbacher, S., Sulem, P., Magnusson, G., Gudjonsson, S.A., Sigurdsson, A., Jonasdottir, As., Jonasdottir, Ad., Wong, W.S.W., Sigurdsson, G., Walters, G.B., Steinberg, S., Helgason, H., Thorleifsson, G., Gudbjartsson, D.F., Helgason, A., Magnusson, O.Th., Thorsteinsdottir. U., & Stefansson, K., (2012) Rate of de novo mutations and the importance of father’s age to disease risk. Nature 488: 471-475.

Sun, J.X., Helgason, A., Masson, S., Ebenesersdóttir, S.S., Li, H., Mallick, S., Gnerre, S., Patterson, N., Kong, A., Reich, R., & Stefansson, K. (2012) A direct characterization of human mutation based on microsatellites. Nature Genetics [press]