Wednesday, June 07, 2017


Although symptoms of autism don't usually appear until a child's second year, and most autistic children aren't diagnosed until they are three or older, recent research shows that the brains of 6-month-old infants who will go on to develop autism already differ from those of normal infants.

These findings may lead to earlier diagnosis and more effective treatment for children who are at risk of developing autism.

They also provide more evidence, if that's still needed, that autism is not caused by childhood vaccinations. As Robert Emerson, the study's lead author points out, "If these differences are already present at six months of age, they would represent a biological foundation for autism that is in place before several vaccines on the CDC schedule that are administered after six months of age, including the Measles, Mumps, Rubella (MMR) inoculation, which is typically given at one year."

A child showing autistic symptoms

Emerson and colleagues at the University of North Carolina at Chapel Hill used functional magnetic resonance imaging (fMRI) while children slept to track brain size, growth and connectivity in 59 infants who were at high risk for autism because they had autistic older siblings. Twenty percent of them could be expected to develop autism, compared to 1.5 percent of children without autistic siblings.

They found a variety of differences between the brains of infants who went on to develop autism and those who didn't.

In keeping with earlier research, those infants who eventually showed autistic symptoms had a faster rate of growth in brain volume and brain surface area between six months and two years of age

The researchers then used advanced artificial intelligence techniques to differentiate between the brains of those infants who did or did not develop autism. Machine-learning programs trained themselves on the brain scans, and were eventually able to identify correctly 82 percent of the children--9 of 11--who would become autistic and 100 percent of those who developed normally.

This very high rate of discrimination was based on nearly 1000 "functional correlations"--how separate regions of the brain connect and work together--that differed between infants who were on autistic versus normal developmental paths.

Schematics representing brain scan signatures at six months that predicted later autism diagnosis in infants. Red bars indicate weaker connections in autistic infant brains, blue bars stronger connections.
[Credit: R.W. Emerson et al., Science Translational Medicine (2017)]

The authors caution that these are preliminary results using state-of-the-art technology, so further research and the development of simpler and less expensive brain-scanning techniques are needed before they can be applied clinically. 

"If future studies confirm these results, detecting brain differences may enable physicians to diagnose and treat autism earlier than they do today," says Diana Bianchi, M.D., Director or the National Institute of Child Health and Development (NICHD).

According to the Centers for Disease Control (CDC), one child out of every 68 in the US will be diagnosed with autism. The earlier that those children can be correctly diagnosed and start to receive treatment--while the brain is most malleable--the better their outcomes.

You can find a summary of an earlier Nature article about this research here, and a link to the abstract of the current study, in Science Translational Medicine, at this URL.


If you enjoyed this post, please subscribe or sign up for email alerts (above right) and let your friends know about

Post a Comment