Several people have inquired about a new paper tying abnormalities in the placenta with an autism outcome in the offspring. This work involved a collaboration between Cheryl Walker and colleagues at UC Davis and Harvey Kliman at Yale in which they analyzed 217 placentas. Those placentas that came from families at high risk for having an autistic child were more likely to have abnormal cell growths called trophoblast inclusions. The at-risk placentas had as many as 15 inclusions while control placentas had no more than 2 such abnormalities. At-risk here means that the family already had a child with autism, which increases the risk 9-fold for the second child to have autism, although that risk is still relatively small (2-7% have autism). Unfortunately, it will be at least a year before we find out how many of the placentas with high numbers of inclusions actually gave rise to autistic children. Kliman had already published a much smaller study in 2006 that indicated that children with autism were 3 times more likely than controls to have inclusions in their placentas.
Elaine Hsiao and I recently reviewed the literature on the placenta and mental disorders, and discussed findings that the placenta is immune activated by maternal infection/immune activation in mouse models. This activation results in altered endocrine function in the placenta. In our own work, we did not see any gross morphological changes in the placenta, however. Surprisingly, Fatemi and colleagues did report morphological changes in the placenta following maternal flu infection in mice. Histology changes in the placenta were also reported following LPS challenge (to mimic bacterial infection) during pregnancy. The most intriguing part of the Yale findings in the human study is that maternal infection would have only occurred in a minority of those cases, so the placental changes must have been caused by other factors, and so placental pathology may be a common mechanism leading to an autistic outcome. It would be really important if maternal infection is tapping into a general mechanism for increasing ASD risk. It may be, of course, that the inclusions are a sign of general placental stress or immune activation, and could be caused by a variety of stressful conditions, and would therefore not be specific to an autism outcome. It will be interesting to determine which types of conditions lead to the formation of inclusions.