Skip navigation

Calls to Action

Can EEG Identify Which High Risk Baby Sibs Need Intervention?

Study of infant brain activity in families affected by autism finds signs of “miswiring” that may help identify babies at highest risk
May 15, 2014

Harvard researchers studying babies in families affected by autism say they’ve found two distinctive types of electroencephalogram (EEG) readings that, together, may pinpoint which of these infants are at the highest risk of developing the disorder. They are presenting their findings this week at the International Meeting for Autism Research (IMFAR), in Atlanta.

See all our IMFAR 2014 news and perspective here.

The Harvard team is part of Autism Speaks Baby Siblings Research Consortium. Previous research by other consortium members has shown that, when one child in a family is diagnosed with autism, the chances that a younger sibling will develop the disorder are around 1 in 5. Still needed are reliable ways to determine which of these 1 in 5 “baby sibs” will go on to develop autism. Such measures would allow intensive early intervention therapies to be focused on the young children who truly need them.

“EEG evaluations could become an important part of such assessments,” comments Paul Wang, Autism Speaks senior vice president and head of medical research. “Studies such as this Harvard investigation also advance our understanding of the brain physiology of autism.”

Electroencephalography is a noninvasive method for recording brain activity both within and across brain regions. EEG readings are of particular interest in autism, as many investigators believe its symptoms result from brain circuit “miswiring.” 

“Using different analytic routines and elements of EEG, we were able to consistently distinguish high risk from low risk infants," said Charles Nelson, the study's senior investigator. "In some cases, we were able to distinguish which high-risk infants would subsequently develop autism from those who would not.”

More specifically, the team found differences in the EEG "gamma band" that distinguished high-risk from low-risk infants. An additional sign - the lack of "mu suppression" - predicted more strongly which high-risk babies would, in fact, develop autism. The researchers propose that the two measures, used together, may be particularly useful in flagging babies at particularly high risk of developing autism. The study involved 208 infants ages 3 to 36 months old.

Co-investigator April Levin described the team's findings at the IMFAR press conference. View her full remarks in the video clip above.