F. Xavier Castellanos, M.D., a professor and director of the The Phyllis Green and Randolph Cowen Institute for Pediatric Neuroscience at the NYU Child Study Center in New York City, has been exploring the use of methods that assess patterns of spontaneous fluctuations in the brain. He thinks that they might hold the key to identifying "fingerprints" of the functioning brain in disorders like ADHD and autism. More importantly, these methods may also provide clinically useful measures for assessing response to treatment. Dr. Castellanos is using these methods as part of a new study funded by the High Risk, High Impact (HRHI) initiative from Autism Speaks titled Early Development of Brain Connectivity in Autism.
Why is this study important?
One of the current hypotheses about the biological basis of autism is that individuals with autism differ in their patterns of brain connectivity. Specifically, it's believed that individuals with autism may have decreased amounts of long-distance brain connections. Because conversation between distant brain regions is necessary for the most complex and integrative brain functions, including emotion and behavior, the "functional underconnectivity" hypothesis of autism may explain the social, communication, and behavioral challenges seen in autism. However, although it is one of the favored hypotheses regarding what may generate autism, technological and experimental limits have made it difficult to test this theory in a sufficiently large number of subjects to be confident about the results.
The new methods that Dr. Castellanos and his team bring from his experience in studying ADHD, schizophrenia and depression may just provide the Rosetta Stone for understanding brain connections in autism. First, by removing the usual requirements for imaging studies that children focus their attention, perform a task, or even stay awake, the number of children who can participate is greatly increased. This approach also makes it possible to study much younger children than before. In fact, in Dr. Castellanos' HRHI study, participants are young—between 3 and 5 years old—and they are studied during natural sleep. Second, the study has been streamlined so that once the child is sleeping the entire experiment can be done in only 30 minutes. Because this is the first study of this type, some of the components of the 30-minute scan are for validation purposes. If successful, future studies can be even faster, permitting this sort of detailed assessment of brain connections in many more children.
What do we know about brain connections during rest or sleep?
The brain never rests. Brain cells always communicate with their "teammates" even when they don't have something new to say. It's a little like chatting about the weather – we do it to be friendly; brain cells do it to keep their connections in good shape, like players on a team. Brain "teams" are essentially brain circuits. The "players" (brain cells) that communicate a lot with each other invest in better connections. It turns out that by measuring spontaneous fluctuations in brain activity, even during sleep, scientists can pick out dozens of brain "teams" because each brain circuit has a slightly different rhythm and location. Some of these circuits are involved in allowing us to see; others to move our muscles; but others help us to understand what people are thinking or feeling. One very special "team"—called the default mode network—sets the level of brain activity when the body is not active with external matters such as seeing, listening, or moving around. The default mode network is concerned with internally-directed activity, like thinking and feeling. Dr. Castellanos' group, led by Michael P. Milham, M.D., Ph.D., has recently published an elegant series of 15 scientific papers demonstrating how much information about brain circuitry can be obtained by measuring spontaneous brain activity—that is activity measured in the resting or sleeping brain. Importantly, the analysis of these circuits revealed that the default mode network had causal effects. For example, a strong and active default mode network might influence the brain circuits that control how one sees, but the circuits that control how one sees do not appear to influence activity in the default network (PubMed Link: 18219617).
One of the tricks in this research is choosing the right starting point from the more than 100,000 possibilities. Adriana Di Martino, M.D., along with her colleagues at the NYU Child Study Center examined previous autism fMRI studies and found that activity in two specific areas were lower in children with autism as compared with typical children (PubMed Link: 18996505). These areas are the anterior and posterior cingulate cortex and together they are involved in processing sensory and motor events and their emotional content, especially the anterior cingulate cortex. These cortical brain regions (see image below of brain with the two regions glowing yellow) represent "nodes" in the default-mode network of brain activity.
The yellow areas on the image of the brain mark the anterior (left) and posterior (right) cingulate cortex, which is less active in individuals with autism than it is in typically-developed individuals. These areas form part of the brain's default-mode network.
These nodes on the default-mode network are working together coherently in adults, but not yet in children. When two nodes are "functionally connected" their activities are synchronized and communication between the two nodes can be effective. Dr. Castellanos and his group have recently shown that as children age into adulthood, the length of connections in the default-mode network change. Connections are "mostly short" during childhood, but many more long connections emerge with adulthood. In terms of functional connectivity, that means that nearby nodes are communicating effectively in children, but distant nodes are much less synchronized. Exciting preliminary data from school-age children with autism show that the length of these default-mode network connections are shorter than in typical peers. That is, the functional connectivity of nodes in the default-mode network in children with autism resembles a more "immature" state of neurodevelopment.
In August, the Castellanos group published a study that highlights another intriguing piece of the autism puzzle using functional connectivity. The paper showed how variation in the default-mode network, centered on the anterior cingulate cortex node, correlates with measures of "autistic traits" in a typical adult population. (PubMed Link: 19605539) The Social Responsiveness Scale for Adults (SRS-A) has been validated as a tool for measuring autistic traits, especially those related to social functioning. The paper showed that the degree of social competence, as rated by a person's significant other, was related to how well-connected their anterior cingulate cortex was with another important brain region known as the insula.
Taken together, these results build a compelling story for investigating the functional connectivity of the brain during sleep in young children with autism, and that is just what Dr. Castellanos and his colleagues will be doing with their new study. They are currently recruiting 3-5 year old children with autism, children experiencing developmental delay without autism spectrum disorders and typically-developing children for this important imaging study. Sleep and autism do not always make great bedfellows, but these researchers have invested much in preparing the children for the unique sights and sounds of the imaging center and are expert at getting the best from each volunteer. So, if you are interested in getting more than just sleep from your child's nighttime, and you are in the greater New York area, please consider contacting the center about participating in this important study.
For more information about participating in the study, click here.
Read the lay-abstract about the funded HRHI research here.