Wednesday, February 10, 2010

On Autism And Parental Age



Did you know the old received wisdom about autism and parental age? Apparently it was that older fathers tended to sire more autistic children.

I did not know this, even though I follow gender-related popularizations like a hawk. This suggests to me that the popularizers didn't think such a finding was of any wider interest, though it could also be that I just missed the intensive popularizations which followed such a finding (Older Fathers Sire Autistic Children! Alarm!).

Now I do know about this received wisdom, because it has been replaced by one having to do with the mother's age:

Older mothers are far more likely to have autistic children than those who give birth younger, say researchers who have examined data on every birth in California in the 1990s.

The team found that mothers over 40 were 51% more likely to have an autistic child than mothers between 25 and 29 (see table for absolute risks).

"This study challenges a current theory in autism epidemiology that identifies the father's age as a key factor in increasing the risk of having a child with autism," says study author Janie Shelton, of the University of California, Davis (press release).

Well, it seems still to be true that older fathers have some effect on the probability of siring an autistic child:

Older mothers are more likely than younger ones to have a child with autism, and older fathers significantly contribute to the risk of the disorder when their partners are under 30, researchers are reporting.

...

The new findings appeared to question the conclusions of earlier research suggesting that the risk of autism spectrum disorders increased with advancing paternal age, but not with advancing maternal age.

One such study analyzed a large Israeli military database and found that children of fathers 40 or older were more than five times as likely to have an autism disorder as those whose fathers were under 30.

An author of that study, Dr. Dolores Malaspina, a psychiatrist at New York University Langone Medical Center, said Monday that mothers and fathers were usually so close in age that small statistical differences could appear to shift the effect of advanced age from one parent to another.

"It's important we not turn around and blame mothers," Dr. Malaspina said. "The evidence is very, very strong that there is a paternal age effect."

But blaming mothers is fun! That's one thing I have learned while blogging on feminism.

More seriously the parental ages are among those variables which are likely to exhibit multicollinearity. This makes it rather difficult to disentangle the effect of the mother's age from the father's age in general, though this new study may have succeeded in that. I doubt that the data set had very many parents where the mother was in her forties and the father in his twenties, though.