Inside Autism

Why do people with autism struggle with decision making? Psychology’s Gabriela Rosenblau is looking deep inside the brain for answers that may lead to improved treatments for autism and other disorders.

A friend who is new in town asks you for a restaurant recommendation. How do you decide where to send her? Maybe you know something about her dining preferences—Italian or Mexican, formal or casual. Or maybe you’ve shared a meal with her before. But your choice is also based on information you don’t even realize you’re considering: her age, social status, perhaps her culture. And if your friend returns with a bad review, when she asks­ again—if she asks again—you’ll adjust your expectations.

It’s a process called social learning, and it essentially explains how we learn from other people through their examples, and how we learn about others by observing our environment. Most of it is implicit—we are acquiring information without even knowing we are doing it. And we are continually updating our ideas with each new data point we collect.

But a person with autism finds this fundamental social mechanism nearly impossible to perform. “People with autism face unique challenges in social decision making,” explained Gabriela Rosenblau, an assistant professor of cognitive neuroscience in the Columbian College of Arts and Science’s Department of Psychological and Brain Sciences. “They have difficulty attaining the body of knowledge needed to judge a person’s preferences and then adjusting those preferences based on new information.”

But why exactly they struggle with social learning has never been definitively determined. Rosenblau is looking for behavioral and neurological answers to that puzzle. With two recent grants—$1.6 million from the National Institutes of Health (NIH) and $500,000 from the Simons Foundation—she’s studying neural development in young people with autism to determine how they acquire and process social knowledge. Her work delves deeply into the brain with neuroimaging and eye-tracking technology, while she and her student researchers also lead young people through social learning tasks.

“I feel strongly that we can use these results to understand neural development and how the brain underlies complex decision making in a social context—while also improving treatment of autism and other disorders,” Rosenblau said.

“We can use these results to understand neural development and how the brain underlies complex decision making…while also improving treatment of autism.”
— Gabriela Rosenblau

Decisions and Predictions

Since joining GW as a postdoctoral student in 2016, Rosenblau has specialized in understanding decision making from both a social context and the underlying brain mechanisms. Much of her research revolves around studying brain development to determine how disorders such as autism hinder decision making. In the restaurant scenario, for example, most people draw on their implicit knowledge of a person’s likes. If their recommendation falls flat, they adjust it according to the other person’s preferences—a process known as the “prediction error.”

But people with autism struggle with the prediction error. Instead, Rosenblau said, they use their own preferences to predict other people’s likes, dislikes and behavior. “If I know someone dislikes apples, I won’t recommend a similar fruit, say, a peach. But if a person with autism likes peaches himself, he will continue to recommend peaches.”

Rosenblau believes that, during adolescence, the brain of a person with autism encodes its own preference more strongly than others. Through her ongoing projects with young people, Rosenblau and her team are examining different areas of the brain. For the youngest participants, they are focusing on the prefrontal cortex, the region most associated with complex cognitive behavior like personality expression and decision making. For older children, they are targeting the cerebellum, which has traditionally been associated with motor function but is now considered crucial to social development. Rosenblau believes the cerebellum may hold the keys to the prediction error. “The cerebellum is a kind of a blueprint for action, a blueprint for learning,” Rosenblau said.

For her research, Rosenblau collaborates with colleagues in the GW Autism and Neurodevelopmental Disorders Institute, the GW School of Nursing and scholars within her own department, including Professor of Cognitive Neuroscience Sarah Shomstein. Senior biology major and research assistant Samantha Metzger said the information they have compiled by guiding young people through exercises has built a foundation for their ongoing work. “These families came in voluntarily to encourage our research so that one day it would help people with autism gain better resources,” she said. “We’re the scientists, but they’re the heroes.”

Main photo: Chynna Golding (left), a student research assistant in Rosenblau’s lab, demonstrates how to fit a volunteer for an EEG cap at the GW Autism and Neurodevelopmental Institute.